# Models at a glance
<a name="model-cards"></a>

Amazon Bedrock provides access to **104** foundation models from **17** providers. Each model card below includes the following details to help you evaluate and integrate a model into your application:
+ **Model details** — Input/output modalities, supported APIs, and endpoints.
+ **Capabilities and features** — Context window, and supported Amazon Bedrock features.
+ **Pricing** — Link to the Amazon Bedrock pricing page.
+ **Programmatic access** — Model IDs and endpoint URLs for API calls.
+ **Service tiers** — Availability across Standard, Priority, Flex, and Reserved tiers.
+ **Regional availability** — AWS Regions where the model is available and supported inference options.
+ **Quotas and limits** — Default requests-per-minute and tokens-per-minute limits by Region.
+ **Sample code** — Ready-to-run code examples for calling the model.

For cross-model comparisons, see [API compatibility](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html), [Endpoint availability](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html), and [Regional availability](https://docs.aws.amazon.com/bedrock/latest/userguide/models-regions.html).

**Jump to a provider:** [AI21 Labs](model-cards-ai21-labs.md) \$1 [Amazon](model-cards-amazon.md) \$1 [Anthropic](model-cards-anthropic.md) \$1 [Cohere](model-cards-cohere.md) \$1 [DeepSeek](model-cards-deepseek.md) \$1 [Google](model-cards-google.md) \$1 [Meta](model-cards-meta.md) \$1 [MiniMax](model-cards-minimax.md) \$1 [Mistral AI](model-cards-mistral-ai.md) \$1 [Moonshot AI](model-cards-moonshot-ai.md) \$1 [NVIDIA](model-cards-nvidia.md) \$1 [OpenAI](model-cards-openai.md) \$1 [Qwen](model-cards-qwen.md) \$1 [Stability AI](model-cards-stability-ai.md) \$1 [TwelveLabs](model-cards-twelvelabs.md) \$1 [Writer](model-cards-writer.md) \$1 [Z.AI](model-cards-zai.md)

# AI21 Labs
<a name="model-cards-ai21-labs"></a>

The following AI21 Labs models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Jamba 1.5 Large](model-card-ai21-labs-jamba-1-5-large.md) | Jamba 1.5 Large is AI21 Labs' hybrid SSM-Transformer model with 398B total parameters and a 256K context window, optimized for complex reasoning tasks across long documents. | 
| [Jamba 1.5 Mini](model-card-ai21-labs-jamba-1-5-mini.md) | Jamba 1.5 Mini is AI21 Labs' lightweight hybrid SSM-Transformer model with 52B total parameters and a 256K context window, optimized for low-latency enterprise tasks. | 

# Jamba 1.5 Large
<a name="model-card-ai21-labs-jamba-1-5-large"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/ai21.png) AI21 Labs — Jamba 1.5 Large
<a name="model-card-ai21-labs-jamba-1-5-large-header"></a>

## Model Details
<a name="model-card-ai21-labs-jamba-1-5-large-details"></a>

Jamba 1.5 Large is AI21 Labs' hybrid SSM-Transformer model with 398B total parameters and a 256K context window, optimized for complex reasoning tasks across long documents. For more information about model development and performance, see the [model/service card](https://www.ai21.com/blog/announcing-jamba-model-family).
+ **Model launch date:** Aug 22, 2024
+ **Model EOL date:** No sooner than 9/23/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 256K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Mar 2024
+ **Marketplace product ID:** `prod-evcp4w4lurj26`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-ai21-labs-jamba-1-5-large-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-ai21-labs-jamba-1-5-large-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | ai21.jamba-1-5-large-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-ai21-labs-jamba-1-5-large-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-ai21-labs-jamba-1-5-large-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-ai21-labs-jamba-1-5-large-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-ai21-labs-jamba-1-5-large-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='ai21.jamba-1-5-large-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='ai21.jamba-1-5-large-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Jamba 1.5 Mini
<a name="model-card-ai21-labs-jamba-1-5-mini"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/ai21.png) AI21 Labs — Jamba 1.5 Mini
<a name="model-card-ai21-labs-jamba-1-5-mini-header"></a>

## Model Details
<a name="model-card-ai21-labs-jamba-1-5-mini-details"></a>

Jamba 1.5 Mini is AI21 Labs' lightweight hybrid SSM-Transformer model with 52B total parameters and a 256K context window, optimized for low-latency enterprise tasks. For more information about model development and performance, see the [model/service card](https://www.ai21.com/blog/announcing-jamba-model-family).
+ **Model launch date:** Aug 22, 2024
+ **Model EOL date:** No sooner than 9/23/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 256K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Mar 2024
+ **Marketplace product ID:** `prod-ggrzjm65qmjhm`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-ai21-labs-jamba-1-5-mini-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-ai21-labs-jamba-1-5-mini-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | ai21.jamba-1-5-mini-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-ai21-labs-jamba-1-5-mini-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-ai21-labs-jamba-1-5-mini-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-ai21-labs-jamba-1-5-mini-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-ai21-labs-jamba-1-5-mini-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='ai21.jamba-1-5-mini-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='ai21.jamba-1-5-mini-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Amazon
<a name="model-cards-amazon"></a>

The following Amazon models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Nova 2 Lite](model-card-amazon-nova-2-lite.md) | Nova 2 Lite is Amazon's cost-efficient multimodal model for simple automation, document processing, and customer support across text, images, and video. | 
| [Nova 2 Sonic](model-card-amazon-nova-2-sonic.md) | Nova 2 Sonic is Amazon's speech-to-speech foundation model for building natural, real-time voice conversation applications. | 
| [Nova Premier](model-card-amazon-nova-premier.md) | Nova Premier is Amazon's multimodal model for complex reasoning, agentic workflows, and model distillation. | 
| [Nova Sonic](model-card-amazon-nova-sonic.md) | Nova Sonic is Amazon's speech-to-speech model that enables natural, real-time voice conversations with low latency and support for multiple languages. | 
| [Nova Lite](model-card-amazon-nova-lite.md) | Nova Lite is Amazon's low-cost multimodal model that processes text, images, and video inputs for tasks like document analysis and visual Q&A. | 
| [Nova Micro](model-card-amazon-nova-micro.md) | Nova Micro is Amazon's fastest text-only model, optimized for speed and low cost in tasks like summarization, translation, and classification. | 
| [Nova Pro](model-card-amazon-nova-pro.md) | Nova Pro is Amazon's balanced multimodal model offering strong accuracy, speed, and cost for a wide range of tasks across text, images, and video. | 
| [Amazon Nova Multimodal Embeddings](model-card-amazon-amazon-nova-multimodal-embeddings.md) | Amazon Nova Multimodal Embeddings is Amazon's embedding model that converts text, images, and video into vector representations for search and retrieval use cases. | 
| [Nova Canvas](model-card-amazon-nova-canvas.md) | Nova Canvas is Amazon's image generation model that creates studio-quality images from text and image prompts with built-in controls for watermarking and content moderation. | 
| [Nova Reel](model-card-amazon-nova-reel.md) | Nova Reel is Amazon's video generation model that creates short videos from text and image prompts with camera motion controls. | 
| [Titan Text Embeddings V2](model-card-amazon-titan-text-embeddings-v2.md) | Titan Text Embeddings V2 is Amazon's second-generation text embeddings model with configurable output dimensions and improved accuracy for retrieval tasks. | 
| [Titan Image Generator G1 v2](model-card-amazon-titan-image-generator-g1-v2.md) | Titan Image Generator G1 v2 is Amazon's image generation model that creates and edits realistic images from text prompts with built-in watermarking. | 
| [Titan Multimodal Embeddings G1](model-card-amazon-titan-multimodal-embeddings-g1.md) | Titan Multimodal Embeddings G1 is Amazon's model that generates embeddings from text and images for multimodal search and recommendation use cases. | 
| [Titan Embeddings G1 - Text](model-card-amazon-titan-embeddings-g1---text.md) | Titan Text Embeddings G1 is Amazon's text embeddings model that converts text into numerical vector representations for search, personalization, and clustering. | 
| [Titan Text Large](model-card-amazon-titan-text-large.md) | Titan Text Large is Amazon's general-purpose text generation model for tasks like summarization, text generation, and conversational chat. | 
| [Titan Text Embeddings v2](model-card-amazon-titan-text-embeddings-v2-2.md) |  | 

# Nova 2 Lite
<a name="model-card-amazon-nova-2-lite"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova 2 Lite
<a name="model-card-amazon-nova-2-lite-header"></a>

## Model Details
<a name="model-card-amazon-nova-2-lite-details"></a>

Nova 2 Lite is Amazon's cost-efficient multimodal model for simple automation, document processing, and customer support across text, images, and video. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/pdfs/ai/responsible-ai/nova-2-lite/nova-2-lite.pdf).
+ **Model launch date:** Dec 02, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 64K
+ **Knowledge cutoff:** Oct 2025


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-amazon-nova-2-lite-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-lite.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-lite.html)  | 

## Pricing
<a name="model-card-amazon-nova-2-lite-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-2-lite-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-2-lite-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.amazon.nova-2-lite-v1:0``eu.amazon.nova-2-lite-v1:0` | global.amazon.nova-2-lite-v1:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-2-lite-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-2-lite-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.amazon.nova-2-lite-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-west-1 (Calgary) | ca-west-1 (Calgary), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.amazon.nova-2-lite-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.amazon.nova-2-lite-v1:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-lite.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-lite.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-lite.html)  | 

## Quotas and Limits
<a name="model-card-amazon-nova-2-lite-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-2-lite-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.nova-2-lite-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='amazon.nova-2-lite-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Nova 2 Sonic
<a name="model-card-amazon-nova-2-sonic"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova 2 Sonic
<a name="model-card-amazon-nova-2-sonic-header"></a>

## Model Details
<a name="model-card-amazon-nova-2-sonic-details"></a>

Nova 2 Sonic is Amazon's speech-to-speech foundation model for building natural, real-time voice conversation applications. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-2-sonic/overview.html).
+ **Model launch date:** Dec 2, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 64K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) InvokeModelWithBidirectionalStream |  | 

## Capabilities and Features
<a name="model-card-amazon-nova-2-sonic-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-sonic.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-2-sonic.html)  | 

## Pricing
<a name="model-card-amazon-nova-2-sonic-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-2-sonic-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-2-sonic-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-2-sonic-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-2-sonic-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-nova-2-sonic-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-2-sonic-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:**

```
pip install boto3
```

**Step 4 - Set environment variables:**

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** This model uses `InvokeModelWithBidirectionalStream`. See [API reference](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InvokeModelWithBidirectionalStream.html) for details.

# Nova Premier
<a name="model-card-amazon-nova-premier"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Premier
<a name="model-card-amazon-nova-premier-header"></a>

## Model Details
<a name="model-card-amazon-nova-premier-details"></a>

Nova Premier is Amazon's multimodal model for complex reasoning, agentic workflows, and model distillation. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Oct 31, 2025
+ **Model EOL date:** September 14, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 1M tokens
+ **Max output tokens:** 25K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Oct 2024


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-nova-premier-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-premier-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-premier-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.amazon.nova-premier-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-premier-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-premier-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.amazon.nova-premier-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-amazon-nova-premier-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-premier-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.nova-premier-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='amazon.nova-premier-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Nova Sonic
<a name="model-card-amazon-nova-sonic"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Sonic
<a name="model-card-amazon-nova-sonic-header"></a>

## Model Details
<a name="model-card-amazon-nova-sonic-details"></a>

Nova Sonic is Amazon's speech-to-speech model that enables natural, real-time voice conversations with low latency and support for multiple languages. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Mar 2025
+ **Model EOL date:** September 14, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) InvokeModelWithBidirectionalStream |  | 

## Pricing
<a name="model-card-amazon-nova-sonic-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-sonic-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-sonic-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-sonic-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-sonic-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-nova-sonic-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-sonic-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:**

```
pip install boto3
```

**Step 4 - Set environment variables:**

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** This model uses `InvokeModelWithBidirectionalStream`. See [API reference](https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InvokeModelWithBidirectionalStream.html) for details.

# Nova Lite
<a name="model-card-amazon-nova-lite"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Lite
<a name="model-card-amazon-nova-lite-header"></a>

## Model Details
<a name="model-card-amazon-nova-lite-details"></a>

Nova Lite is Amazon's low-cost multimodal model that processes text, images, and video inputs for tasks like document analysis and visual Q&A. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Dec 05, 2024
+ **Model EOL date:** No sooner than 12/4/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 300K tokens
+ **Max output tokens:** 5K
+ **Knowledge cutoff:** Oct 2024


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-nova-lite-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-lite-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-lite-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.amazon.nova-lite-v1:0``eu.amazon.nova-lite-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-lite-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-lite-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| me-central-1 (UAE) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.amazon.nova-lite-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.amazon.nova-lite-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| il-central-1 (Tel Aviv) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-west-1 (Ireland), eu-west-3 (Paris), il-central-1 (Tel Aviv) | 

## Quotas and Limits
<a name="model-card-amazon-nova-lite-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-lite-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.nova-lite-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='amazon.nova-lite-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Nova Micro
<a name="model-card-amazon-nova-micro"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Micro
<a name="model-card-amazon-nova-micro-header"></a>

## Model Details
<a name="model-card-amazon-nova-micro-details"></a>

Nova Micro is Amazon's fastest text-only model, optimized for speed and low cost in tasks like summarization, translation, and classification. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Dec 05, 2024
+ **Model EOL date:** No sooner than 12/4/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 5K
+ **Knowledge cutoff:** Oct 2024


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-nova-micro-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-micro-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-micro-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.amazon.nova-micro-v1:0``eu.amazon.nova-micro-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-micro-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-micro-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.amazon.nova-micro-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.amazon.nova-micro-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| il-central-1 (Tel Aviv) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-west-1 (Ireland), eu-west-3 (Paris), il-central-1 (Tel Aviv) | 

## Quotas and Limits
<a name="model-card-amazon-nova-micro-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-micro-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.nova-micro-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='amazon.nova-micro-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Nova Pro
<a name="model-card-amazon-nova-pro"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Pro
<a name="model-card-amazon-nova-pro-header"></a>

## Model Details
<a name="model-card-amazon-nova-pro-details"></a>

Nova Pro is Amazon's balanced multimodal model offering strong accuracy, speed, and cost for a wide range of tasks across text, images, and video. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Dec 05, 2024
+ **Model EOL date:** No sooner than 12/4/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 300K tokens
+ **Max output tokens:** 5K
+ **Knowledge cutoff:** Oct 2024


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-amazon-nova-pro-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-pro.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-nova-pro.html)  | 

## Pricing
<a name="model-card-amazon-nova-pro-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-pro-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-pro-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.amazon.nova-pro-v1:0``eu.amazon.nova-pro-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-pro-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-pro-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| me-central-1 (UAE) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.amazon.nova-pro-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.amazon.nova-pro-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| il-central-1 (Tel Aviv) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-west-1 (Ireland), eu-west-3 (Paris), il-central-1 (Tel Aviv) | 

## Quotas and Limits
<a name="model-card-amazon-nova-pro-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-pro-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.nova-pro-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='amazon.nova-pro-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Amazon Nova Multimodal Embeddings
<a name="model-card-amazon-amazon-nova-multimodal-embeddings"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Amazon Nova Multimodal Embeddings
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-header"></a>

## Model Details
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-details"></a>

Amazon Nova Multimodal Embeddings is Amazon's embedding model that converts text, images, and video into vector representations for search and retrieval use cases. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Dec 3, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) StartAsyncInvoke |  | 

## Capabilities and Features
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  —  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-amazon-amazon-nova-multimodal-embeddings.html)  | 

## Pricing
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-2-multimodal-embeddings-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-amazon-nova-multimodal-embeddings-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:**

```
pip install boto3
```

**Step 4 - Set environment variables:**

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** This model uses `StartAsyncInvoke`. Save the file as `bedrock-first-request.py`

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.start_async_invoke(
    modelId='amazon.nova-2-multimodal-embeddings-v1:0',
    modelInput={},
    outputDataConfig={'s3OutputDataConfig': {'s3Uri': 's3://your-bucket/output/'}}
)
print(response)
```

# Nova Canvas
<a name="model-card-amazon-nova-canvas"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Canvas
<a name="model-card-amazon-nova-canvas-header"></a>

## Model Details
<a name="model-card-amazon-nova-canvas-details"></a>

Nova Canvas is Amazon's image generation model that creates studio-quality images from text and image prompts with built-in controls for watermarking and content moderation. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Dec 3, 2024
+ **Model EOL date:** September 30, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy (certain regions)


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-nova-canvas-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-canvas-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-canvas-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-canvas-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-canvas-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-nova-canvas-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-canvas-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.nova-canvas-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Nova Reel
<a name="model-card-amazon-nova-reel"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Nova Reel
<a name="model-card-amazon-nova-reel-header"></a>

## Model Details
<a name="model-card-amazon-nova-reel-details"></a>

Nova Reel is Amazon's video generation model that creates short videos from text and image prompts with camera motion controls. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html).
+ **Model launch date:** Dec 3, 2024
+ **Model EOL date:** September 30, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy (certain regions)


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) StartAsyncInvoke |  | 

## Pricing
<a name="model-card-amazon-nova-reel-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-nova-reel-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.nova-reel-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-nova-reel-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-nova-reel-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-nova-reel-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-nova-reel-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:**

```
pip install boto3
```

**Step 4 - Set environment variables:**

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** This model uses `StartAsyncInvoke`. Save the file as `bedrock-first-request.py`

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.start_async_invoke(
    modelId='amazon.nova-reel-v1:0',
    modelInput={},
    outputDataConfig={'s3OutputDataConfig': {'s3Uri': 's3://your-bucket/output/'}}
)
print(response)
```

# Titan Text Embeddings V2
<a name="model-card-amazon-titan-text-embeddings-v2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Titan Text Embeddings V2
<a name="model-card-amazon-titan-text-embeddings-v2-header"></a>

## Model Details
<a name="model-card-amazon-titan-text-embeddings-v2-details"></a>

Titan Text Embeddings V2 is Amazon's second-generation text embeddings model with configurable output dimensions and improved accuracy for retrieval tasks. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/titan-text-embeddings/overview.html).
+ **Model launch date:** Apr 30, 2024
+ **Model EOL date:** No sooner than 4/30/2024
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 8K tokens


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-titan-text-embeddings-v2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-titan-text-embeddings-v2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.titan-embed-text-v2:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-titan-text-embeddings-v2-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-titan-text-embeddings-v2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-east-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-2 (Zurich) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-2 (Spain) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-2 (Seoul) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-3 (Osaka) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-2 (Hyderabad) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-titan-text-embeddings-v2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-titan-text-embeddings-v2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.titan-embed-text-v2:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Titan Image Generator G1 v2
<a name="model-card-amazon-titan-image-generator-g1-v2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Titan Image Generator G1 v2
<a name="model-card-amazon-titan-image-generator-g1-v2-header"></a>

## Model Details
<a name="model-card-amazon-titan-image-generator-g1-v2-details"></a>

Titan Image Generator G1 v2 is Amazon's image generation model that creates and edits realistic images from text prompts with built-in watermarking. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/titan-text-embeddings/overview.html).
+ **Model launch date:** Nov 29, 2023
+ **Model EOL date:** June 30, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-titan-image-generator-g1-v2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-titan-image-generator-g1-v2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.titan-image-generator-v2:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-titan-image-generator-g1-v2-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-titan-image-generator-g1-v2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-titan-image-generator-g1-v2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-titan-image-generator-g1-v2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.titan-image-generator-v2:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Titan Multimodal Embeddings G1
<a name="model-card-amazon-titan-multimodal-embeddings-g1"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Titan Multimodal Embeddings G1
<a name="model-card-amazon-titan-multimodal-embeddings-g1-header"></a>

## Model Details
<a name="model-card-amazon-titan-multimodal-embeddings-g1-details"></a>

Titan Multimodal Embeddings G1 is Amazon's model that generates embeddings from text and images for multimodal search and recommendation use cases. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/titan-text-embeddings/overview.html).
+ **Model launch date:** Nov 29, 2023
+ **Model EOL date:** No sooner than 11/29/2024
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-titan-multimodal-embeddings-g1-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-titan-multimodal-embeddings-g1-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.titan-embed-image-v1 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-titan-multimodal-embeddings-g1-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-titan-multimodal-embeddings-g1-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-titan-multimodal-embeddings-g1-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-titan-multimodal-embeddings-g1-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.titan-embed-image-v1',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Titan Embeddings G1 - Text
<a name="model-card-amazon-titan-embeddings-g1---text"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Titan Embeddings G1 - Text
<a name="model-card-amazon-titan-embeddings-g1---text-header"></a>

## Model Details
<a name="model-card-amazon-titan-embeddings-g1---text-details"></a>

Titan Text Embeddings G1 is Amazon's text embeddings model that converts text into numerical vector representations for search, personalization, and clustering. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/titan-text-embeddings/overview.html).
+ **Model launch date:** Sep 28, 2023
+ **Model EOL date:** No sooner than 11/29/2024
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 8K tokens


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-titan-embeddings-g1---text-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-titan-embeddings-g1---text-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.titan-embed-text-v1 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-titan-embeddings-g1---text-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-titan-embeddings-g1---text-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-titan-embeddings-g1---text-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-titan-embeddings-g1---text-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='amazon.titan-embed-text-v1',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Titan Text Large
<a name="model-card-amazon-titan-text-large"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Titan Text Large
<a name="model-card-amazon-titan-text-large-header"></a>

## Model Details
<a name="model-card-amazon-titan-text-large-details"></a>

Titan Text Large is Amazon's general-purpose text generation model for tasks like summarization, text generation, and conversational chat. For more information about model development and performance, see the [model/service card](https://docs.aws.amazon.com/ai/responsible-ai/titan-text-embeddings/overview.html).
+ **Model launch date:** Sep 28, 2023
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 8K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-titan-text-large-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-titan-text-large-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.titan-tg1-large | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-amazon-titan-text-large-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-amazon-titan-text-large-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-titan-text-large-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-titan-text-large-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='amazon.titan-tg1-large',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Titan Text Embeddings v2
<a name="model-card-amazon-titan-text-embeddings-v2-2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nova2pro.png) Amazon — Titan Text Embeddings v2
<a name="model-card-amazon-titan-text-embeddings-v2-2-header"></a>

## Model Details
<a name="model-card-amazon-titan-text-embeddings-v2-2-details"></a>

Model description coming soon.
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 8K tokens


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-amazon-titan-text-embeddings-v2-2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-amazon-titan-text-embeddings-v2-2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | amazon.titan-embed-g1-text-02 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Regional Availability
<a name="model-card-amazon-titan-text-embeddings-v2-2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-amazon-titan-text-embeddings-v2-2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-amazon-titan-text-embeddings-v2-2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

# Anthropic
<a name="model-cards-anthropic"></a>

The following Anthropic models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Claude Opus 4.7](model-card-anthropic-claude-opus-4-7.md) | Claude Opus 4.7 is Anthropic's most capable generally available model, advancing performance across coding, enterprise workflows, and long-running agentic tasks. | 
| [Claude Mythos Preview](model-card-anthropic-claude-mythos-preview.md) | Claude Mythos Preview (gated research preview) is a new class of intelligence built for ambitious projects, and the world's best model for cybersecurity, autonomous coding, and long-running agents. | 
| [Claude Sonnet 4.6](model-card-anthropic-claude-sonnet-4-6.md) | Claude Sonnet 4.6 is a full upgrade of Anthropic's mid-tier model with improved coding, computer use, long-context reasoning, and agent planning with a 1M token context window. | 
| [Claude Opus 4.6](model-card-anthropic-claude-opus-4-6.md) | Claude Opus 4.6 is Anthropic's flagship model that plans more carefully, sustains agentic tasks longer, and operates reliably in massive codebases. | 
| [Claude Opus 4.5](model-card-anthropic-claude-opus-4-5.md) | Claude Opus 4.5 is Anthropic's model for coding, agents, and computer use with improvements for spreadsheets and long-running chats. | 
| [Claude Haiku 4.5](model-card-anthropic-claude-haiku-4-5.md) | Claude Haiku 4.5 is Anthropic's lightweight model optimized for speed and efficiency with strong coding and agent performance. | 
| [Claude Sonnet 4.5](model-card-anthropic-claude-sonnet-4-5.md) | Claude Sonnet 4.5 is Anthropic's model optimized for agents, coding, and computer use with significant improvements across all benchmarks. | 
| [Claude Opus 4.1](model-card-anthropic-claude-opus-4-1.md) | Claude Opus 4.1 is an upgrade to Anthropic's model with improved coding, reasoning, and agentic task capabilities. | 
| [Claude Sonnet 4](model-card-anthropic-claude-sonnet-4.md) | Claude Sonnet 4 is Anthropic's balanced model with strong coding and reasoning capabilities, improved instruction following, and extended thinking with tool use. | 
| [Claude 3.5 Haiku](model-card-anthropic-claude-3-5-haiku.md) | Claude 3.5 Haiku is Anthropic's next-generation fast model with improved coding and reasoning performance over Claude 3 Haiku at the same speed tier. | 
| [Claude 3 Haiku](model-card-anthropic-claude-3-haiku.md) | Claude 3 Haiku is Anthropic's fastest and most compact Claude 3 model, optimized for speed and efficiency in near-instant responses. | 

# Claude Opus 4.7
<a name="model-card-anthropic-claude-opus-4-7"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Opus 4.7
<a name="model-card-anthropic-claude-opus-4-7-header"></a>

## Model Details
<a name="model-card-anthropic-claude-opus-4-7-details"></a>

Claude Opus 4.7 is Anthropic's most capable generally available model, advancing performance across coding, enterprise workflows, and long-running agentic tasks.
+ **Model launch date:** Apr 16, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 128K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** January 2026
+ **Marketplace product ID:** `prod-d2ik6zgct5hxi`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Messages |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-opus-4-7-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  | 

## Pricing
<a name="model-card-anthropic-claude-opus-4-7-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-opus-4-7-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-opus-4-7 | N/A | `us.anthropic.claude-opus-4-7``eu.anthropic.claude-opus-4-7``jp.anthropic.claude-opus-4-7` | global.anthropic.claude-opus-4-7 | 
| bedrock-mantle | anthropic.claude-opus-4-7 | https://bedrock-mantle.\$1region\$1.api.aws/anthropic/v1/messages | N/A | N/A | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/anthropic/v1/messages".*

## Service Tiers
<a name="model-card-anthropic-claude-opus-4-7-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-opus-4-7-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-opus-4-7`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-west-1 (Calgary) | ca-west-1 (Calgary), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-opus-4-7`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

**Geo: JP**

Geo Inference ID: `jp.anthropic.claude-opus-4-7`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| ap-northeast-1 (Tokyo) | ap-northeast-1 (Tokyo), ap-northeast-3 (Osaka) | 
| ap-northeast-3 (Osaka) | ap-northeast-1 (Tokyo), ap-northeast-3 (Osaka) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-opus-4-7 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-7.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-opus-4-7-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

Default quotas for Claude Opus 4.7 are 10M TPM on `bedrock-mantle` and 15M TPM on `bedrock-runtime` for each supported region.

## Sampling parameters no longer supported
<a name="model-card-anthropic-claude-opus-4-7-parameters-change"></a>

Starting with Claude Opus 4.7, `temperature`, `top_p`, and `top_k` parameters are no longer supported. The recommended migration path is to omit these parameters entirely from your requests and to use prompting to guide the model's behavior. If you previously relied on temperature = 0 for deterministic outputs, note that this setting did not guarantee identical responses across invocations.

## Sample Code
<a name="model-card-anthropic-claude-opus-4-7-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Messages API ]

```
pip install -U "anthropic[bedrock]"
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Messages API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Messages API ]

```
from anthropic import AnthropicBedrockMantle

client = AnthropicBedrockMantle(aws_region="us-east-1")

message = client.messages.create(
    model="anthropic.claude-opus-4-7",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}],
)

print(message.content[0].text)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-opus-4-7',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-opus-4-7',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Mythos Preview
<a name="model-card-anthropic-claude-mythos-preview"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Mythos Preview
<a name="model-card-anthropic-claude-mythos-preview-header"></a>

## Model Details
<a name="model-card-anthropic-claude-mythos-preview-details"></a>

According to Anthropic, Claude Mythos Preview (gated research preview) is a new class of intelligence built for ambitious projects focusing on cybersecurity, autonomous coding, and long-running agents.

Available only as a gated research preview with access prioritized for defensive cybersecurity use cases.
+ **Model launch date:** Apr 07, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Preview
+ **Context window:** 1M tokens
+ **Max output tokens:** 128K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Dec 2025


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Messages |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-mythos-preview-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-mythos-preview.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-mythos-preview.html)  | 

## Pricing
<a name="model-card-anthropic-claude-mythos-preview-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-mythos-preview-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-mantle | anthropic.claude-mythos-preview | https://bedrock-mantle.\$1region\$1.api.aws/v1 | N/A | N/A | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-mantle endpoint URL will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-mythos-preview-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-mythos-preview-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

# Claude Sonnet 4.6
<a name="model-card-anthropic-claude-sonnet-4-6"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Sonnet 4.6
<a name="model-card-anthropic-claude-sonnet-4-6-header"></a>

## Model Details
<a name="model-card-anthropic-claude-sonnet-4-6-details"></a>

Claude Sonnet 4.6 is a full upgrade of Anthropic's mid-tier model with improved coding, computer use, long-context reasoning, and agent planning with a 1M token context window. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/bbd8ef16d70b7a1665f14f306ee88b53f686aa75.pdf).
+ **Model launch date:** Feb 17, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 64K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Aug 2025
+ **Marketplace product ID:** `prod-ffvjxvh4ltq64`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-sonnet-4-6-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-6.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-6.html)  | 

## Pricing
<a name="model-card-anthropic-claude-sonnet-4-6-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-sonnet-4-6-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-sonnet-4-6 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-sonnet-4-6``eu.anthropic.claude-sonnet-4-6``au.anthropic.claude-sonnet-4-6` | global.anthropic.claude-sonnet-4-6 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-sonnet-4-6-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-sonnet-4-6-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-sonnet-4-6`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-west-1 (Calgary) | ca-west-1 (Calgary), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-sonnet-4-6`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

**Geo: AU**

Geo Inference ID: `au.anthropic.claude-sonnet-4-6`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| ap-southeast-2 (Sydney) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-4 (Melbourne) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-6 (New Zealand) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne), ap-southeast-6 (New Zealand) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-sonnet-4-6 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-6.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-6.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-6.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-sonnet-4-6-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-sonnet-4-6-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-sonnet-4-6',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-sonnet-4-6',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Opus 4.6
<a name="model-card-anthropic-claude-opus-4-6"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Opus 4.6
<a name="model-card-anthropic-claude-opus-4-6-header"></a>

## Model Details
<a name="model-card-anthropic-claude-opus-4-6-details"></a>

Claude Opus 4.6 is Anthropic's flagship model that plans more carefully, sustains agentic tasks longer, and operates reliably in massive codebases. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/0dd865075ad3132672ee0ab40b05a53f14cf5288.pdf).
+ **Model launch date:** Feb 5, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 128K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** May 2025
+ **Marketplace product ID:** `prod-5ukwuglpt66kg`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-opus-4-6-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-6.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-6.html)  | 

## Pricing
<a name="model-card-anthropic-claude-opus-4-6-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-opus-4-6-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-opus-4-6-v1 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-opus-4-6-v1``eu.anthropic.claude-opus-4-6-v1``au.anthropic.claude-opus-4-6-v1` | global.anthropic.claude-opus-4-6-v1 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-opus-4-6-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-opus-4-6-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-opus-4-6-v1`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-west-1 (Calgary) | ca-west-1 (Calgary), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-opus-4-6-v1`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

**Geo: AU**

Geo Inference ID: `au.anthropic.claude-opus-4-6-v1`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| ap-southeast-2 (Sydney) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-4 (Melbourne) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-6 (New Zealand) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne), ap-southeast-6 (New Zealand) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-opus-4-6-v1 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-6.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-6.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-6.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-opus-4-6-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-opus-4-6-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-opus-4-6-v1',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-opus-4-6-v1',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Opus 4.5
<a name="model-card-anthropic-claude-opus-4-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Opus 4.5
<a name="model-card-anthropic-claude-opus-4-5-header"></a>

## Model Details
<a name="model-card-anthropic-claude-opus-4-5-details"></a>

Claude Opus 4.5 is Anthropic's model for coding, agents, and computer use with improvements for spreadsheets and long-running chats. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/bf10f64990cfda0ba858290be7b8cc6317685f47.pdf).
+ **Model launch date:** Nov 24, 2025
+ **Model EOL date:** No sooner than 3/25/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 200K tokens
+ **Max output tokens:** 64K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Mar 2025
+ **Marketplace product ID:** `prod-jhuafngbly644`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-opus-4-5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-5.html)  | 

## Pricing
<a name="model-card-anthropic-claude-opus-4-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-opus-4-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-opus-4-5-20251101-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-opus-4-5-20251101-v1:0``eu.anthropic.claude-opus-4-5-20251101-v1:0` | global.anthropic.claude-opus-4-5-20251101-v1:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-opus-4-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-opus-4-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-opus-4-5-20251101-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-opus-4-5-20251101-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-opus-4-5-20251101-v1:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-5.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-opus-4-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-opus-4-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-opus-4-5-20251101-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-opus-4-5-20251101-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Haiku 4.5
<a name="model-card-anthropic-claude-haiku-4-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Haiku 4.5
<a name="model-card-anthropic-claude-haiku-4-5-header"></a>

## Model Details
<a name="model-card-anthropic-claude-haiku-4-5-details"></a>

Claude Haiku 4.5 is Anthropic's lightweight model optimized for speed and efficiency with strong coding and agent performance. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/7aad69bf12627d42234e01ee7c36305dc2f6a970.pdf).
+ **Model launch date:** Oct 16, 2025
+ **Model EOL date:** No sooner than 10/1/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 200K tokens
+ **Max output tokens:** 64K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Feb 2025
+ **Marketplace product ID:** `prod-xdkflymybwmvi`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-haiku-4-5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-haiku-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-haiku-4-5.html)  | 

## Pricing
<a name="model-card-anthropic-claude-haiku-4-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-haiku-4-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-haiku-4-5-20251001-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-haiku-4-5-20251001-v1:0``eu.anthropic.claude-haiku-4-5-20251001-v1:0``au.anthropic.claude-haiku-4-5-20251001-v1:0` | global.anthropic.claude-haiku-4-5-20251001-v1:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-haiku-4-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-haiku-4-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-haiku-4-5-20251001-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-haiku-4-5-20251001-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

**Geo: AU**

Geo Inference ID: `au.anthropic.claude-haiku-4-5-20251001-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| ap-southeast-2 (Sydney) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-4 (Melbourne) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-6 (New Zealand) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne), ap-southeast-6 (New Zealand) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-haiku-4-5-20251001-v1:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-haiku-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-haiku-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-haiku-4-5.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-haiku-4-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-haiku-4-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-haiku-4-5-20251001-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-haiku-4-5-20251001-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Sonnet 4.5
<a name="model-card-anthropic-claude-sonnet-4-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Sonnet 4.5
<a name="model-card-anthropic-claude-sonnet-4-5-header"></a>

## Model Details
<a name="model-card-anthropic-claude-sonnet-4-5-details"></a>

Claude Sonnet 4.5 is Anthropic's model optimized for agents, coding, and computer use with significant improvements across all benchmarks. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/963373e433e489a87a10c823c52a0a013e9172dd.pdf).
+ **Model launch date:** Sep 30, 2025
+ **Model EOL date:** No sooner than 9/29/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 200K tokens
+ **Max output tokens:** 64K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Apr 2025
+ **Marketplace product ID:** `prod-mxcfnwvpd6kb4`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-sonnet-4-5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-5.html)  | 

## Pricing
<a name="model-card-anthropic-claude-sonnet-4-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-sonnet-4-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-sonnet-4-5-20250929-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-sonnet-4-5-20250929-v1:0``eu.anthropic.claude-sonnet-4-5-20250929-v1:0``au.anthropic.claude-sonnet-4-5-20250929-v1:0` | global.anthropic.claude-sonnet-4-5-20250929-v1:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-sonnet-4-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-sonnet-4-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-gov-east-1 (GovCloud) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-6 (New Zealand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-sonnet-4-5-20250929-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-gov-east-1 (AWS GovCloud (US-East)) | us-gov-east-1 (AWS GovCloud (US-East)), us-gov-west-1 (AWS GovCloud (US-West)) | 
| us-gov-west-1 (AWS GovCloud (US-West)) | us-gov-east-1 (AWS GovCloud (US-East)), us-gov-west-1 (AWS GovCloud (US-West)) | 
| ca-central-1 (Canada) | ca-central-1 (Canada), us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-sonnet-4-5-20250929-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

**Geo: AU**

Geo Inference ID: `au.anthropic.claude-sonnet-4-5-20250929-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| ap-southeast-2 (Sydney) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-4 (Melbourne) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne) | 
| ap-southeast-6 (New Zealand) | ap-southeast-2 (Sydney), ap-southeast-4 (Melbourne), ap-southeast-6 (New Zealand) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-sonnet-4-5-20250929-v1:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4-5.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-sonnet-4-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-sonnet-4-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-sonnet-4-5-20250929-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-sonnet-4-5-20250929-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Opus 4.1
<a name="model-card-anthropic-claude-opus-4-1"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Opus 4.1
<a name="model-card-anthropic-claude-opus-4-1-header"></a>

## Model Details
<a name="model-card-anthropic-claude-opus-4-1-details"></a>

Claude Opus 4.1 is an upgrade to Anthropic's model with improved coding, reasoning, and agentic task capabilities. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/9fa30625273bafdf5af82c93719d7ca606485a16.pdf).
+ **Model launch date:** Aug 05, 2025
+ **Model EOL date:** May 31, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 200K tokens
+ **Max output tokens:** 64K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Mar 2025
+ **Marketplace product ID:** `prod-w3q2d6rfge4tw`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-opus-4-1-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-1.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-opus-4-1.html)  | 

## Pricing
<a name="model-card-anthropic-claude-opus-4-1-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-opus-4-1-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-opus-4-1-20250805-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.anthropic.claude-opus-4-1-20250805-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-opus-4-1-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-opus-4-1-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-opus-4-1-20250805-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-anthropic-claude-opus-4-1-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-opus-4-1-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-opus-4-1-20250805-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-opus-4-1-20250805-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude Sonnet 4
<a name="model-card-anthropic-claude-sonnet-4"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude Sonnet 4
<a name="model-card-anthropic-claude-sonnet-4-header"></a>

## Model Details
<a name="model-card-anthropic-claude-sonnet-4-details"></a>

Claude Sonnet 4 is Anthropic's balanced model with strong coding and reasoning capabilities, improved instruction following, and extended thinking with tool use. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/4263b940cabb546aa0e3283f35b686f4f3b2ff47.pdf).
+ **Model launch date:** May 23, 2025
+ **Model EOL date:** October 14, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy (certain regions)
+ **Context window:** 200K tokens
+ **Max output tokens:** 64K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Mar 2025
+ **Marketplace product ID:** `prod-4pmewlybdftbs`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-anthropic-claude-sonnet-4-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4.html)  | 

## Pricing
<a name="model-card-anthropic-claude-sonnet-4-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-sonnet-4-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-sonnet-4-20250514-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-sonnet-4-20250514-v1:0``eu.anthropic.claude-sonnet-4-20250514-v1:0` | global.anthropic.claude-sonnet-4-20250514-v1:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-sonnet-4-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-sonnet-4-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-sonnet-4-20250514-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-sonnet-4-20250514-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| il-central-1 (Tel Aviv) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris), il-central-1 (Tel Aviv) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.anthropic.claude-sonnet-4-20250514-v1:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-anthropic-claude-sonnet-4.html)  | 

## Quotas and Limits
<a name="model-card-anthropic-claude-sonnet-4-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-sonnet-4-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-sonnet-4-20250514-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-sonnet-4-20250514-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude 3.5 Haiku
<a name="model-card-anthropic-claude-3-5-haiku"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude 3.5 Haiku
<a name="model-card-anthropic-claude-3-5-haiku-header"></a>

## Model Details
<a name="model-card-anthropic-claude-3-5-haiku-details"></a>

Claude 3.5 Haiku is Anthropic's next-generation fast model with improved coding and reasoning performance over Claude 3 Haiku at the same speed tier. For more information about model development and performance, see the [model/service card](https://assets.anthropic.com/m/1cd9d098ac3e6467/original/Claude-3-Model-Card-October-Addendum.pdf).
+ **Model launch date:** Nov 4, 2024
+ **Model EOL date:** June 19, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 200K tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Jul 2024
+ **Marketplace product ID:** `prod-5oba7y7jpji56`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-anthropic-claude-3-5-haiku-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-3-5-haiku-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-3-5-haiku-20241022-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.anthropic.claude-3-5-haiku-20241022-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-3-5-haiku-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-3-5-haiku-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-3-5-haiku-20241022-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-anthropic-claude-3-5-haiku-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-3-5-haiku-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-3-5-haiku-20241022-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-3-5-haiku-20241022-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Claude 3 Haiku
<a name="model-card-anthropic-claude-3-haiku"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/claude.png) Anthropic — Claude 3 Haiku
<a name="model-card-anthropic-claude-3-haiku-header"></a>

## Model Details
<a name="model-card-anthropic-claude-3-haiku-details"></a>

Claude 3 Haiku is Anthropic's fastest and most compact Claude 3 model, optimized for speed and efficiency in near-instant responses. For more information about model development and performance, see the [model/service card](https://www-cdn.anthropic.com/de8ba9b01c9ab7cbabf5c33b80b7bbc618857627/Model_Card_Claude_3.pdf).
+ **Model launch date:** Mar 13, 2024
+ **Model EOL date:** September 10, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy (certain regions)
+ **Context window:** 200K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Aug 2023
+ **Marketplace product ID:** `prod-ozonys2hmmpeu`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-anthropic-claude-3-haiku-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-anthropic-claude-3-haiku-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | anthropic.claude-3-haiku-20240307-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.anthropic.claude-3-haiku-20240307-v1:0``eu.anthropic.claude-3-haiku-20240307-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-anthropic-claude-3-haiku-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-anthropic-claude-3-haiku-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-east-1 (GovCloud) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-2 (Zurich) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-2 (Seoul) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-1 (Singapore) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.anthropic.claude-3-haiku-20240307-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 
| us-gov-east-1 (AWS GovCloud (US-East)) | us-gov-east-1 (AWS GovCloud (US-East)), us-gov-west-1 (AWS GovCloud (US-West)) | 

**Geo: EU**

Geo Inference ID: `eu.anthropic.claude-3-haiku-20240307-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 

## Quotas and Limits
<a name="model-card-anthropic-claude-3-haiku-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-anthropic-claude-3-haiku-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='anthropic.claude-3-haiku-20240307-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='anthropic.claude-3-haiku-20240307-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Cohere
<a name="model-cards-cohere"></a>

The following Cohere models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Embed v4](model-card-cohere-embed-v4.md) | Embed v4 is Cohere's unified multimodal embedding model that processes text, images, and mixed content in a single model for search and RAG. | 
| [Rerank 3.5](model-card-cohere-rerank-3-5.md) | Rerank 3.5 is Cohere's relevance scoring model that reorders search results for improved accuracy in RAG pipelines and enterprise search. | 
| [Command R](model-card-cohere-command-r.md) | Command R is Cohere's scalable LLM optimized for retrieval-augmented generation and tool use in enterprise applications with a 128K context window. | 
| [Command R\$1](model-card-cohere-command-r-plus.md) | Command R\$1 is Cohere's model for complex RAG workflows, multi-step tool use, and enterprise tasks with a 128K context window. | 
| [Embed English](model-card-cohere-embed-english.md) | Embed English is Cohere's English-language text embedding model for search, classification, and clustering with strong retrieval accuracy. | 
| [Embed Multilingual](model-card-cohere-embed-multilingual.md) | Embed Multilingual is Cohere's multilingual text embedding model supporting 100\$1 languages for cross-lingual search and classification. | 

# Embed v4
<a name="model-card-cohere-embed-v4"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/cohere.png) Cohere — Embed v4
<a name="model-card-cohere-embed-v4-header"></a>

## Model Details
<a name="model-card-cohere-embed-v4-details"></a>

Embed v4 is Cohere's unified multimodal embedding model that processes text, images, and mixed content in a single model for search and RAG. For more information about model development and performance, see the [model/service card](https://docs.cohere.com/docs/cohere-embed).
+ **Model launch date:** Apr 15, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Marketplace product ID:** `prod-ft3cj5gst3spo`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-cohere-embed-v4-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-cohere-embed-v4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-cohere-embed-v4.html)  | 

## Pricing
<a name="model-card-cohere-embed-v4-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-cohere-embed-v4-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | cohere.embed-v4:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.cohere.embed-v4:0``eu.cohere.embed-v4:0` | global.cohere.embed-v4:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-cohere-embed-v4-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-cohere-embed-v4-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.cohere.embed-v4:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.cohere.embed-v4:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.cohere.embed-v4:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-cohere-embed-v4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-cohere-embed-v4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-cohere-embed-v4.html)  | 

## Quotas and Limits
<a name="model-card-cohere-embed-v4-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-cohere-embed-v4-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='cohere.embed-v4:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Rerank 3.5
<a name="model-card-cohere-rerank-3-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/cohere.png) Cohere — Rerank 3.5
<a name="model-card-cohere-rerank-3-5-header"></a>

## Model Details
<a name="model-card-cohere-rerank-3-5-details"></a>

Rerank 3.5 is Cohere's relevance scoring model that reorders search results for improved accuracy in RAG pipelines and enterprise search. For more information about model development and performance, see the [model/service card](https://docs.cohere.com/docs/rerank).
+ **Model launch date:** Dec 2, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 4K tokens
+ **Marketplace product ID:** `prod-2o5bej62oxkbi`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-cohere-rerank-3-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-cohere-rerank-3-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | cohere.rerank-v3-5:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-cohere-rerank-3-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-cohere-rerank-3-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-cohere-rerank-3-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-cohere-rerank-3-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='cohere.rerank-v3-5:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Command R
<a name="model-card-cohere-command-r"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/cohere.png) Cohere — Command R
<a name="model-card-cohere-command-r-header"></a>

## Model Details
<a name="model-card-cohere-command-r-details"></a>

Command R is Cohere's scalable LLM optimized for retrieval-augmented generation and tool use in enterprise applications with a 128K context window. For more information about model development and performance, see the [model/service card](https://docs.cohere.com/docs/command-r).
+ **Model launch date:** Aug 2024
+ **Model EOL date:** August 19, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Mar 2024
+ **Marketplace product ID:** `prod-tukx4z3hrewle`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-cohere-command-r-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-cohere-command-r-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | cohere.command-r-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-cohere-command-r-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-cohere-command-r-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-cohere-command-r-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-cohere-command-r-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='cohere.command-r-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='cohere.command-r-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Command R\$1
<a name="model-card-cohere-command-r-plus"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/cohere.png) Cohere — Command R\$1
<a name="model-card-cohere-command-r-plus-header"></a>

## Model Details
<a name="model-card-cohere-command-r-plus-details"></a>

Command R\$1 is Cohere's model for complex RAG workflows, multi-step tool use, and enterprise tasks with a 128K context window. For more information about model development and performance, see the [model/service card](https://docs.cohere.com/docs/command-r-plus).
+ **Model launch date:** Aug 2024
+ **Model EOL date:** August 19, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Mar 2024
+ **Marketplace product ID:** `prod-nb4wqmplze2pm`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-cohere-command-r-plus-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-cohere-command-r-plus-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | cohere.command-r-plus-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-cohere-command-r-plus-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-cohere-command-r-plus-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-cohere-command-r-plus-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-cohere-command-r-plus-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='cohere.command-r-plus-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='cohere.command-r-plus-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Embed English
<a name="model-card-cohere-embed-english"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/cohere.png) Cohere — Embed English
<a name="model-card-cohere-embed-english-header"></a>

## Model Details
<a name="model-card-cohere-embed-english-details"></a>

Embed English is Cohere's English-language text embedding model for search, classification, and clustering with strong retrieval accuracy. For more information about model development and performance, see the [model/service card](https://docs.cohere.com/docs/cohere-embed).
+ **Model launch date:** Nov 2, 2023
+ **Model EOL date:** No sooner than 11/30/2024
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 512 tokens
+ **Marketplace product ID:** `b7568428-a1ab-46d8-bab3-37def50f6f6a`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-cohere-embed-english-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-cohere-embed-english-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | cohere.embed-english-v3 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-cohere-embed-english-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-cohere-embed-english-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-1 (Singapore) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-cohere-embed-english-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-cohere-embed-english-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='cohere.embed-english-v3',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Embed Multilingual
<a name="model-card-cohere-embed-multilingual"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/cohere.png) Cohere — Embed Multilingual
<a name="model-card-cohere-embed-multilingual-header"></a>

## Model Details
<a name="model-card-cohere-embed-multilingual-details"></a>

Embed Multilingual is Cohere's multilingual text embedding model supporting 100\$1 languages for cross-lingual search and classification. For more information about model development and performance, see the [model/service card](https://docs.cohere.com/docs/cohere-embed).
+ **Model launch date:** Nov 2, 2023
+ **Model EOL date:** No sooner than 11/13/2024
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 512 tokens
+ **Marketplace product ID:** `38e55671-c3fe-4a44-9783-3584906e7cad`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-cohere-embed-multilingual-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-cohere-embed-multilingual-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | cohere.embed-multilingual-v3 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-cohere-embed-multilingual-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-cohere-embed-multilingual-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-1 (Singapore) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-cohere-embed-multilingual-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-cohere-embed-multilingual-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='cohere.embed-multilingual-v3',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# DeepSeek
<a name="model-cards-deepseek"></a>

The following DeepSeek models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [DeepSeek V3.2](model-card-deepseek-deepseek-v3-2.md) | DeepSeek V3.2 is DeepSeek's mixture-of-experts model with improved reasoning, coding, and instruction following capabilities. | 
| [DeepSeek-V3.1](model-card-deepseek-deepseek-v3-1.md) | DeepSeek-V3.1 is DeepSeek's 685B parameter mixture-of-experts model with strong performance across coding, math, and general reasoning tasks. | 
| [DeepSeek-R1](model-card-deepseek-deepseek-r1.md) | DeepSeek-R1 is DeepSeek's reasoning model that uses chain-of-thought to solve complex math, coding, and logic problems. | 

# DeepSeek V3.2
<a name="model-card-deepseek-deepseek-v3-2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/deepseek.png) DeepSeek — DeepSeek V3.2
<a name="model-card-deepseek-deepseek-v3-2-header"></a>

## Model Details
<a name="model-card-deepseek-deepseek-v3-2-details"></a>

DeepSeek V3.2 is DeepSeek's mixture-of-experts model with improved reasoning, coding, and instruction following capabilities. For more information about model development and performance, see the [model/service card](https://api-docs.deepseek.com/news/news251201).
+ **Model launch date:** Dec 01, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Mar 2025


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-deepseek-deepseek-v3-2-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-2.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-2.html)  | 

## Pricing
<a name="model-card-deepseek-deepseek-v3-2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-deepseek-deepseek-v3-2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | deepseek.v3.2 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | deepseek.v3.2 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-deepseek-deepseek-v3-2-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-deepseek-deepseek-v3-2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-deepseek-deepseek-v3-2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-deepseek-deepseek-v3-2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="deepseek.v3.2",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="deepseek.v3.2",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='deepseek.v3.2',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='deepseek.v3.2',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# DeepSeek-V3.1
<a name="model-card-deepseek-deepseek-v3-1"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/deepseek.png) DeepSeek — DeepSeek-V3.1
<a name="model-card-deepseek-deepseek-v3-1-header"></a>

## Model Details
<a name="model-card-deepseek-deepseek-v3-1-details"></a>

DeepSeek-V3.1 is DeepSeek's 685B parameter mixture-of-experts model with strong performance across coding, math, and general reasoning tasks. For more information about model development and performance, see the [model/service card](https://api-docs.deepseek.com/news/news250821).
+ **Model launch date:** Aug 21, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Jan 2025


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-deepseek-deepseek-v3-1-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-1.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-1.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-1.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-v3-1.html)  | 

## Pricing
<a name="model-card-deepseek-deepseek-v3-1-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-deepseek-deepseek-v3-1-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | deepseek.v3-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | deepseek.v3.1 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-deepseek-deepseek-v3-1-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-deepseek-deepseek-v3-1-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-deepseek-deepseek-v3-1-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-deepseek-deepseek-v3-1-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="deepseek.v3.1",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="deepseek.v3.1",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='deepseek.v3-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='deepseek.v3-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# DeepSeek-R1
<a name="model-card-deepseek-deepseek-r1"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/deepseek.png) DeepSeek — DeepSeek-R1
<a name="model-card-deepseek-deepseek-r1-header"></a>

## Model Details
<a name="model-card-deepseek-deepseek-r1-details"></a>

DeepSeek-R1 is DeepSeek's reasoning model that uses chain-of-thought to solve complex math, coding, and logic problems. For more information about model development and performance, see the [model/service card](https://api-docs.deepseek.com/news/news250120).
+ **Model launch date:** Jan 20, 2025
+ **Model EOL date:** No sooner than 3/10/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Reasoning:** Supported
+ **Knowledge cutoff:** Jan 2025


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-deepseek-deepseek-r1-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-r1.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-deepseek-deepseek-r1.html)  | 

## Pricing
<a name="model-card-deepseek-deepseek-r1-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-deepseek-deepseek-r1-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | deepseek.r1-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.deepseek.r1-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-deepseek-deepseek-r1-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-deepseek-deepseek-r1-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.deepseek.r1-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-deepseek-deepseek-r1-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-deepseek-deepseek-r1-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='deepseek.r1-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='deepseek.r1-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Google
<a name="model-cards-google"></a>

The following Google models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Gemma 3 12B IT](model-card-google-gemma-3-12b-it.md) | Gemma 3 12B IT is Google's 12-billion parameter open model with instruction tuning, supporting text and image inputs with a 128K context window. | 
| [Gemma 3 27B PT](model-card-google-gemma-3-27b-pt.md) | Gemma 3 27B PT is Google's 27-billion parameter open model, the largest in the Gemma 3 family, with multimodal capabilities and a 128K context window. | 
| [Gemma 3 4B IT](model-card-google-gemma-3-4b-it.md) | Gemma 3 4B IT is Google's compact 4-billion parameter open model with instruction tuning, designed for on-device and edge deployment. | 

# Gemma 3 12B IT
<a name="model-card-google-gemma-3-12b-it"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/google.png) Google — Gemma 3 12B IT
<a name="model-card-google-gemma-3-12b-it-header"></a>

## Model Details
<a name="model-card-google-gemma-3-12b-it-details"></a>

Gemma 3 12B IT is Google's 12-billion parameter open model with instruction tuning, supporting text and image inputs with a 128K context window. For more information about model development and performance, see the [model/service card](https://ai.google.dev/gemma/docs/core/model_card_3).
+ **Model launch date:** Mar 12, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-google-gemma-3-12b-it-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-12b-it.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-12b-it.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-12b-it.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-12b-it.html)  | 

## Pricing
<a name="model-card-google-gemma-3-12b-it-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-google-gemma-3-12b-it-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | google.gemma-3-12b-it | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | google.gemma-3-12b-it | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-google-gemma-3-12b-it-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-google-gemma-3-12b-it-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-google-gemma-3-12b-it-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-google-gemma-3-12b-it-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="google.gemma-3-12b-it",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="google.gemma-3-12b-it",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='google.gemma-3-12b-it',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='google.gemma-3-12b-it',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Gemma 3 27B PT
<a name="model-card-google-gemma-3-27b-pt"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/google.png) Google — Gemma 3 27B PT
<a name="model-card-google-gemma-3-27b-pt-header"></a>

## Model Details
<a name="model-card-google-gemma-3-27b-pt-details"></a>

Gemma 3 27B PT is Google's 27-billion parameter open model, the largest in the Gemma 3 family, with multimodal capabilities and a 128K context window. For more information about model development and performance, see the [model/service card](https://ai.google.dev/gemma/docs/core/model_card_3).
+ **Model launch date:** Mar 12, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-google-gemma-3-27b-pt-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-27b-pt.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-27b-pt.html)  | 

## Pricing
<a name="model-card-google-gemma-3-27b-pt-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-google-gemma-3-27b-pt-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | google.gemma-3-27b-it | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-google-gemma-3-27b-pt-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-google-gemma-3-27b-pt-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-google-gemma-3-27b-pt-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-google-gemma-3-27b-pt-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-runtime.<your-region>.amazonaws.com/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="google.gemma-3-27b-it",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='google.gemma-3-27b-it',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='google.gemma-3-27b-it',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Gemma 3 4B IT
<a name="model-card-google-gemma-3-4b-it"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/google.png) Google — Gemma 3 4B IT
<a name="model-card-google-gemma-3-4b-it-header"></a>

## Model Details
<a name="model-card-google-gemma-3-4b-it-details"></a>

Gemma 3 4B IT is Google's compact 4-billion parameter open model with instruction tuning, designed for on-device and edge deployment. For more information about model development and performance, see the [model/service card](https://ai.google.dev/gemma/docs/core/model_card_3).
+ **Model launch date:** Mar 12, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-google-gemma-3-4b-it-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-4b-it.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-4b-it.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-4b-it.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-google-gemma-3-4b-it.html)  | 

## Pricing
<a name="model-card-google-gemma-3-4b-it-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-google-gemma-3-4b-it-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | google.gemma-3-4b-it | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | google.gemma-3-4b-it | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-google-gemma-3-4b-it-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-google-gemma-3-4b-it-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-google-gemma-3-4b-it-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-google-gemma-3-4b-it-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="google.gemma-3-4b-it",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="google.gemma-3-4b-it",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='google.gemma-3-4b-it',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='google.gemma-3-4b-it',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Meta
<a name="model-cards-meta"></a>

The following Meta models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Llama 4 Maverick 17B Instruct](model-card-meta-llama-4-maverick-17b-instruct.md) | Llama 4 Maverick is Meta's 17-billion active parameter mixture-of-experts model with 128 experts, optimized for multimodal chat and instruction following. | 
| [Llama 4 Scout 17B Instruct](model-card-meta-llama-4-scout-17b-instruct.md) | Llama 4 Scout is Meta's 17-billion active parameter mixture-of-experts model with 16 experts and a 10M token context window for long-document tasks. | 
| [Llama 3.3 70B Instruct](model-card-meta-llama-3-3-70b-instruct.md) | Llama 3.3 70B Instruct is Meta's 70-billion parameter model with improved efficiency, delivering strong reasoning and coding performance with a 128K context window. | 
| [Llama 3.2 11B Instruct](model-card-meta-llama-3-2-11b-instruct.md) | Llama 3.2 11B Instruct is Meta's 11-billion parameter multimodal model that processes both text and images with a 128K context window. | 
| [Llama 3.2 1B Instruct](model-card-meta-llama-3-2-1b-instruct.md) | Llama 3.2 1B Instruct is Meta's ultra-lightweight 1-billion parameter model optimized for on-device and edge deployment. | 
| [Llama 3.2 3B Instruct](model-card-meta-llama-3-2-3b-instruct.md) | Llama 3.2 3B Instruct is Meta's 3-billion parameter lightweight model designed for on-device tasks like summarization and instruction following. | 
| [Llama 3.2 90B Instruct](model-card-meta-llama-3-2-90b-instruct.md) | Llama 3.2 90B Instruct is Meta's 90-billion parameter multimodal model that processes text and images for visual reasoning and document understanding. | 
| [Llama 3.1 405B Instruct](model-card-meta-llama-3-1-405b-instruct.md) | Llama 3.1 405B Instruct is Meta's largest open model with 405 billion parameters and a 128K context window, supporting tool use and multilingual tasks. | 
| [Llama 3.1 70B Instruct](model-card-meta-llama-3-1-70b-instruct.md) | Llama 3.1 70B Instruct is Meta's 70-billion parameter model with an extended 128K context window and support for tool use and code generation. | 
| [Llama 3.1 8B Instruct](model-card-meta-llama-3-1-8b-instruct.md) | Llama 3.1 8B Instruct is Meta's compact 8-billion parameter model with a 128K context window, suitable for edge deployment and fine-tuning. | 
| [Llama 3 70B Instruct](model-card-meta-llama-3-70b-instruct.md) | Llama 3 70B Instruct is Meta's 70-billion parameter instruction-tuned model with an 8K context window, optimized for dialogue and assistant-like tasks. | 
| [Llama 3 8B Instruct](model-card-meta-llama-3-8b-instruct.md) | Llama 3 8B Instruct is Meta's 8-billion parameter instruction-tuned model with an 8K context window, designed for efficient deployment on smaller infrastructure. | 

# Llama 4 Maverick 17B Instruct
<a name="model-card-meta-llama-4-maverick-17b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 4 Maverick 17B Instruct
<a name="model-card-meta-llama-4-maverick-17b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-4-maverick-17b-instruct-details"></a>

Llama 4 Maverick is Meta's 17-billion active parameter mixture-of-experts model with 128 experts, optimized for multimodal chat and instruction following. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama4/).
+ **Model launch date:** Apr 05, 2025
+ **Model EOL date:** No sooner than 4/28/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Aug 2024


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-4-maverick-17b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-4-maverick-17b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama4-maverick-17b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama4-maverick-17b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-4-maverick-17b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-4-maverick-17b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama4-maverick-17b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-4-maverick-17b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-4-maverick-17b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama4-maverick-17b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama4-maverick-17b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 4 Scout 17B Instruct
<a name="model-card-meta-llama-4-scout-17b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 4 Scout 17B Instruct
<a name="model-card-meta-llama-4-scout-17b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-4-scout-17b-instruct-details"></a>

Llama 4 Scout is Meta's 17-billion active parameter mixture-of-experts model with 16 experts and a 10M token context window for long-document tasks. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama4/).
+ **Model launch date:** Apr 05, 2025
+ **Model EOL date:** No sooner than 4/28/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 10M tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Aug 2024


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-4-scout-17b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-4-scout-17b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama4-scout-17b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama4-scout-17b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-4-scout-17b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-4-scout-17b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama4-scout-17b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-4-scout-17b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-4-scout-17b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama4-scout-17b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama4-scout-17b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.3 70B Instruct
<a name="model-card-meta-llama-3-3-70b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.3 70B Instruct
<a name="model-card-meta-llama-3-3-70b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-3-70b-instruct-details"></a>

Llama 3.3 70B Instruct is Meta's 70-billion parameter model with improved efficiency, delivering strong reasoning and coding performance with a 128K context window. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_3/).
+ **Model launch date:** Dec 06, 2024
+ **Model EOL date:** No sooner than 12/19/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-3-70b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-3-70b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-3-70b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama3-3-70b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-3-70b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-3-70b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-3-70b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-3-70b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-3-70b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-3-70b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-3-70b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.2 11B Instruct
<a name="model-card-meta-llama-3-2-11b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.2 11B Instruct
<a name="model-card-meta-llama-3-2-11b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-2-11b-instruct-details"></a>

Llama 3.2 11B Instruct is Meta's 11-billion parameter multimodal model that processes both text and images with a 128K context window. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_2/).
+ **Model launch date:** Sep 25, 2024
+ **Model EOL date:** July 7, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-2-11b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-2-11b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-2-11b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama3-2-11b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-2-11b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-2-11b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-2-11b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-2-11b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-2-11b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-2-11b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-2-11b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.2 1B Instruct
<a name="model-card-meta-llama-3-2-1b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.2 1B Instruct
<a name="model-card-meta-llama-3-2-1b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-2-1b-instruct-details"></a>

Llama 3.2 1B Instruct is Meta's ultra-lightweight 1-billion parameter model optimized for on-device and edge deployment. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_2/).
+ **Model launch date:** Sep 25, 2024
+ **Model EOL date:** July 7, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-2-1b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-2-1b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-2-1b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.meta.llama3-2-1b-instruct-v1:0``eu.meta.llama3-2-1b-instruct-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-2-1b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-2-1b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-2-1b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.meta.llama3-2-1b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-2-1b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-2-1b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-2-1b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-2-1b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.2 3B Instruct
<a name="model-card-meta-llama-3-2-3b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.2 3B Instruct
<a name="model-card-meta-llama-3-2-3b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-2-3b-instruct-details"></a>

Llama 3.2 3B Instruct is Meta's 3-billion parameter lightweight model designed for on-device tasks like summarization and instruction following. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_2/).
+ **Model launch date:** Sep 25, 2024
+ **Model EOL date:** July 7, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-2-3b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-2-3b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-2-3b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.meta.llama3-2-3b-instruct-v1:0``eu.meta.llama3-2-3b-instruct-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-2-3b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-2-3b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-2-3b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.meta.llama3-2-3b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-3 (Paris) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-2-3b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-2-3b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-2-3b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-2-3b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.2 90B Instruct
<a name="model-card-meta-llama-3-2-90b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.2 90B Instruct
<a name="model-card-meta-llama-3-2-90b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-2-90b-instruct-details"></a>

Llama 3.2 90B Instruct is Meta's 90-billion parameter multimodal model that processes text and images for visual reasoning and document understanding. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_2/).
+ **Model launch date:** Sep 25, 2024
+ **Model EOL date:** July 7, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-2-90b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-2-90b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-2-90b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama3-2-90b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-2-90b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-2-90b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-2-90b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-2-90b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-2-90b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-2-90b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-2-90b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.1 405B Instruct
<a name="model-card-meta-llama-3-1-405b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.1 405B Instruct
<a name="model-card-meta-llama-3-1-405b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-1-405b-instruct-details"></a>

Llama 3.1 405B Instruct is Meta's largest open model with 405 billion parameters and a 128K context window, supporting tool use and multilingual tasks. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_1/).
+ **Model launch date:** Jul 23, 2024
+ **Model EOL date:** Legacy: July 7, 2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Legacy
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-1-405b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-1-405b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-1-405b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama3-1-405b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-1-405b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-1-405b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-1-405b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Sample Code
<a name="model-card-meta-llama-3-1-405b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-1-405b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.1 70B Instruct
<a name="model-card-meta-llama-3-1-70b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.1 70B Instruct
<a name="model-card-meta-llama-3-1-70b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-1-70b-instruct-details"></a>

Llama 3.1 70B Instruct is Meta's 70-billion parameter model with an extended 128K context window and support for tool use and code generation. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_1/).
+ **Model launch date:** Jul 23, 2024
+ **Model EOL date:** No sooner than 7/23/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-1-70b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-1-70b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-1-70b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama3-1-70b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-1-70b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-1-70b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-1-70b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-1-70b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-1-70b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-1-70b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-1-70b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3.1 8B Instruct
<a name="model-card-meta-llama-3-1-8b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3.1 8B Instruct
<a name="model-card-meta-llama-3-1-8b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-1-8b-instruct-details"></a>

Llama 3.1 8B Instruct is Meta's compact 8-billion parameter model with a 128K context window, suitable for edge deployment and fine-tuning. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_1/).
+ **Model launch date:** Jul 23, 2024
+ **Model EOL date:** No sooner than 7/23/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-meta-llama-3-1-8b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-1-8b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-1-8b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.meta.llama3-1-8b-instruct-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-1-8b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-1-8b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.meta.llama3-1-8b-instruct-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-1-8b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-1-8b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-1-8b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-1-8b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3 70B Instruct
<a name="model-card-meta-llama-3-70b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3 70B Instruct
<a name="model-card-meta-llama-3-70b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-70b-instruct-details"></a>

Llama 3 70B Instruct is Meta's 70-billion parameter instruction-tuned model with an 8K context window, optimized for dialogue and assistant-like tasks. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/other-models/#meta-llama-3).
+ **Model launch date:** Apr 18, 2024
+ **Model EOL date:** No sooner than 4/23/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 8K tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-meta-llama-3-70b-instruct-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-70b-instruct.html)  | 

## Pricing
<a name="model-card-meta-llama-3-70b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-70b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-70b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-70b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-70b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-70b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-70b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-70b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-70b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Llama 3 8B Instruct
<a name="model-card-meta-llama-3-8b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/meta.png) Meta — Llama 3 8B Instruct
<a name="model-card-meta-llama-3-8b-instruct-header"></a>

## Model Details
<a name="model-card-meta-llama-3-8b-instruct-details"></a>

Llama 3 8B Instruct is Meta's 8-billion parameter instruction-tuned model with an 8K context window, designed for efficient deployment on smaller infrastructure. For more information about model development and performance, see the [model/service card](https://www.llama.com/docs/model-cards-and-prompt-formats/other-models/#meta-llama-3).
+ **Model launch date:** Apr 18, 2024
+ **Model EOL date:** No sooner than 4/23/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 8K tokens
+ **Max output tokens:** 8K
+ **Knowledge cutoff:** Dec 2023


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-meta-llama-3-8b-instruct-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-meta-llama-3-8b-instruct.html)  | 

## Pricing
<a name="model-card-meta-llama-3-8b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-meta-llama-3-8b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | meta.llama3-8b-instruct-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-meta-llama-3-8b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-meta-llama-3-8b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-gov-west-1 (GovCloud) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-meta-llama-3-8b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-meta-llama-3-8b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='meta.llama3-8b-instruct-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='meta.llama3-8b-instruct-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# MiniMax
<a name="model-cards-minimax"></a>

The following MiniMax models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [MiniMax M2](model-card-minimax-minimax-m2.md) | MiniMax M2 is MiniMax's large language model with strong multilingual capabilities and solid performance on reasoning and coding benchmarks. | 
| [MiniMax M2.1](model-card-minimax-minimax-m2-1.md) | MiniMax M2.1 is MiniMax's updated model with improved reasoning, coding, and instruction following over M2. | 
| [MiniMax M2.5](model-card-minimax-minimax-m2-5.md) | MiniMax M2.5 is an agent-native frontier model trained to reason efficiently, decompose tasks optimally, and complete complex workflows under real-world time and cost constraints. | 

# MiniMax M2
<a name="model-card-minimax-minimax-m2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/minimax2.1.png) MiniMax — MiniMax M2
<a name="model-card-minimax-minimax-m2-header"></a>

## Model Details
<a name="model-card-minimax-minimax-m2-details"></a>

MiniMax M2 is MiniMax's large language model with strong multilingual capabilities and solid performance on reasoning and coding benchmarks. For more information about model development and performance, see the [model/service card](https://www.minimax.io/news/minimax-m2).
+ **Model launch date:** Oct 23, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-minimax-minimax-m2-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2.html)  | 

## Pricing
<a name="model-card-minimax-minimax-m2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-minimax-minimax-m2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | minimax.minimax-m2 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | minimax.minimax-m2 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-minimax-minimax-m2-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-minimax-minimax-m2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-minimax-minimax-m2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-minimax-minimax-m2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="minimax.minimax-m2",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="minimax.minimax-m2",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='minimax.minimax-m2',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='minimax.minimax-m2',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# MiniMax M2.1
<a name="model-card-minimax-minimax-m2-1"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/minimax2.1.png) MiniMax — MiniMax M2.1
<a name="model-card-minimax-minimax-m2-1-header"></a>

## Model Details
<a name="model-card-minimax-minimax-m2-1-details"></a>

MiniMax M2.1 is MiniMax's updated model with improved reasoning, coding, and instruction following over M2. For more information about model development and performance, see the [model/service card](https://www.minimax.io/news/minimax-m21).
+ **Model launch date:** Dec 23, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-minimax-minimax-m2-1-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-1.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-1.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-1.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-1.html)  | 

## Pricing
<a name="model-card-minimax-minimax-m2-1-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-minimax-minimax-m2-1-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | minimax.minimax-m2.1 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | minimax.minimax-m2.1 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-minimax-minimax-m2-1-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-minimax-minimax-m2-1-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-minimax-minimax-m2-1-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-minimax-minimax-m2-1-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="minimax.minimax-m2.1",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="minimax.minimax-m2.1",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='minimax.minimax-m2.1',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='minimax.minimax-m2.1',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# MiniMax M2.5
<a name="model-card-minimax-minimax-m2-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/minimax2.1.png) MiniMax — MiniMax M2.5
<a name="model-card-minimax-minimax-m2-5-header"></a>

## Model Details
<a name="model-card-minimax-minimax-m2-5-details"></a>

MiniMax M2.5 is an agent-native frontier model trained explicitly to reason efficiently, decompose tasks optimally, and complete complex workflows under real-world time and cost constraints. It achieves task completion speeds comparable to or faster than leading proprietary frontier models by combining high inference throughput with reinforcement learning focused on token-efficient reasoning and better decision-making in agentic scaffolds. For more information about model development and performance, see the [model/service card](https://minimax.io/news/minimax-m2).
+ **Model launch date:** Feb 12, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 1M tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-minimax-minimax-m2-5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-5.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-minimax-minimax-m2-5.html)  | 

## Pricing
<a name="model-card-minimax-minimax-m2-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-minimax-minimax-m2-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | minimax.minimax-m2.5 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | minimax.minimax-m2.5 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-minimax-minimax-m2-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-minimax-minimax-m2-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-minimax-minimax-m2-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-minimax-minimax-m2-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="minimax.minimax-m2.5",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="minimax.minimax-m2.5",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='minimax.minimax-m2.5',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='minimax.minimax-m2.5',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Mistral AI
<a name="model-cards-mistral-ai"></a>

The following Mistral AI models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Mistral Small](model-card-mistral-ai-mistral-small.md) | Mistral Small is Mistral AI's cost-efficient model optimized for low-latency tasks like classification, translation, and customer support. | 
| [Ministral 14B 3.0](model-card-mistral-ai-ministral-14b-3-0.md) | Ministral 14B 3.0 is Mistral AI's 14-billion parameter edge model optimized for on-device deployment with strong performance on knowledge and reasoning tasks. | 
| [Ministral 3 8B](model-card-mistral-ai-ministral-3-8b.md) | Ministral 3 8B is Mistral AI's 8-billion parameter compact model for edge and mobile deployment with efficient inference. | 
| [Ministral 3B](model-card-mistral-ai-ministral-3b.md) | Ministral 3B is Mistral AI's ultra-compact 3-billion parameter model for on-device tasks requiring minimal compute. | 
| [Mistral Large 3](model-card-mistral-ai-mistral-large-3.md) | Mistral Large 3 is Mistral AI's 675-billion parameter model with strong performance on coding, reasoning, and multilingual tasks. | 
| [Voxtral Small 24B 2507](model-card-mistral-ai-voxtral-small-24b-2507.md) | Voxtral Small 24B is Mistral AI's speech-to-text model with 24 billion parameters for high-accuracy transcription and voice understanding. | 
| [Magistral Small 2509](model-card-mistral-ai-magistral-small-2509.md) | Magistral Small 2509 is Mistral AI's reasoning model that uses chain-of-thought to solve complex math, coding, and logic problems. | 
| [Voxtral Mini 3B 2507](model-card-mistral-ai-voxtral-mini-3b-2507.md) | Voxtral Mini 3B is Mistral AI's compact speech-to-text model for real-time transcription and voice understanding on edge devices. | 
| [Devstral 2 123B](model-card-mistral-ai-devstral-2-123b.md) | Devstral 2 123B is Mistral AI's 123-billion parameter coding model optimized for software engineering tasks including code generation, debugging, and refactoring. | 
| [Pixtral Large](model-card-mistral-ai-pixtral-large.md) | Pixtral Large is Mistral AI's 124-billion parameter multimodal model that processes text and images for visual reasoning and document understanding. | 
| [Mistral Large](model-card-mistral-ai-mistral-large.md) | Mistral Large is Mistral AI's flagship model with strong reasoning, multilingual support, and a 32K context window for complex enterprise tasks. | 
| [Mixtral 8x7B Instruct](model-card-mistral-ai-mixtral-8x7b-instruct.md) | Mixtral 8x7B Instruct is Mistral AI's sparse mixture-of-experts model with 8 experts and 7B parameters each, delivering strong performance at faster inference speeds. | 
| [Mistral 7B Instruct](model-card-mistral-ai-mistral-7b-instruct.md) | Mistral 7B Instruct is Mistral AI's 7-billion parameter instruction-tuned model with grouped-query attention and sliding window attention for efficient long-context inference. | 

# Mistral Small
<a name="model-card-mistral-ai-mistral-small"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Mistral Small
<a name="model-card-mistral-ai-mistral-small-header"></a>

## Model Details
<a name="model-card-mistral-ai-mistral-small-details"></a>

Mistral Small is Mistral AI's cost-efficient model optimized for low-latency tasks like classification, translation, and customer support. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Dec 16, 2025
+ **Model EOL date:** No sooner than 5/24/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 32K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-mistral-ai-mistral-small-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-mistral-small-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.mistral-small-2402-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-mistral-small-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-mistral-small-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-mistral-small-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-mistral-small-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.mistral-small-2402-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.mistral-small-2402-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Ministral 14B 3.0
<a name="model-card-mistral-ai-ministral-14b-3-0"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Ministral 14B 3.0
<a name="model-card-mistral-ai-ministral-14b-3-0-header"></a>

## Model Details
<a name="model-card-mistral-ai-ministral-14b-3-0-details"></a>

Ministral 14B 3.0 is Mistral AI's 14-billion parameter edge model optimized for on-device deployment with strong performance on knowledge and reasoning tasks. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Dec 2, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-ministral-14b-3-0-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-14b-3-0.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-14b-3-0.html)  | 

## Pricing
<a name="model-card-mistral-ai-ministral-14b-3-0-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-ministral-14b-3-0-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.ministral-3-14b-instruct | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-ministral-14b-3-0-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-ministral-14b-3-0-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-ministral-14b-3-0-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-ministral-14b-3-0-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-runtime.<your-region>.amazonaws.com/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.ministral-3-14b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.ministral-3-14b-instruct',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.ministral-3-14b-instruct',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Ministral 3 8B
<a name="model-card-mistral-ai-ministral-3-8b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Ministral 3 8B
<a name="model-card-mistral-ai-ministral-3-8b-header"></a>

## Model Details
<a name="model-card-mistral-ai-ministral-3-8b-details"></a>

Ministral 3 8B is Mistral AI's 8-billion parameter compact model for edge and mobile deployment with efficient inference. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Dec 02, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-ministral-3-8b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-3-8b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-3-8b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-3-8b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-3-8b.html)  | 

## Pricing
<a name="model-card-mistral-ai-ministral-3-8b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-ministral-3-8b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.ministral-3-8b-instruct | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | mistral.ministral-3-8b-instruct | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-ministral-3-8b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-ministral-3-8b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-ministral-3-8b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-ministral-3-8b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.ministral-3-8b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.ministral-3-8b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.ministral-3-8b-instruct',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.ministral-3-8b-instruct',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Ministral 3B
<a name="model-card-mistral-ai-ministral-3b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Ministral 3B
<a name="model-card-mistral-ai-ministral-3b-header"></a>

## Model Details
<a name="model-card-mistral-ai-ministral-3b-details"></a>

Ministral 3B is Mistral AI's ultra-compact 3-billion parameter model for on-device tasks requiring minimal compute. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Dec 2, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-ministral-3b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-3b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-ministral-3b.html)  | 

## Pricing
<a name="model-card-mistral-ai-ministral-3b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-ministral-3b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.ministral-3-3b-instruct | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-ministral-3b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-ministral-3b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-ministral-3b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-ministral-3b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-runtime.<your-region>.amazonaws.com/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.ministral-3-3b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.ministral-3-3b-instruct',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.ministral-3-3b-instruct',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Mistral Large 3
<a name="model-card-mistral-ai-mistral-large-3"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Mistral Large 3
<a name="model-card-mistral-ai-mistral-large-3-header"></a>

## Model Details
<a name="model-card-mistral-ai-mistral-large-3-details"></a>

Mistral Large 3 is Mistral AI's 675-billion parameter model with strong performance on coding, reasoning, and multilingual tasks. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Dec 2, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 32K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-mistral-large-3-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large-3.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large-3.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large-3.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large-3.html)  | 

## Pricing
<a name="model-card-mistral-ai-mistral-large-3-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-mistral-large-3-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.mistral-large-3-675b-instruct | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | mistral.mistral-large-3-675b-instruct | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-mistral-large-3-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-mistral-large-3-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-mistral-large-3-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-mistral-large-3-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.mistral-large-3-675b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.mistral-large-3-675b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.mistral-large-3-675b-instruct',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.mistral-large-3-675b-instruct',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Voxtral Small 24B 2507
<a name="model-card-mistral-ai-voxtral-small-24b-2507"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Voxtral Small 24B 2507
<a name="model-card-mistral-ai-voxtral-small-24b-2507-header"></a>

## Model Details
<a name="model-card-mistral-ai-voxtral-small-24b-2507-details"></a>

Voxtral Small 24B is Mistral AI's speech-to-text model with 24 billion parameters for high-accuracy transcription and voice understanding. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Oct 30, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 32K tokens


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-voxtral-small-24b-2507-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-small-24b-2507.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-small-24b-2507.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-small-24b-2507.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-small-24b-2507.html)  | 

## Pricing
<a name="model-card-mistral-ai-voxtral-small-24b-2507-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-voxtral-small-24b-2507-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.voxtral-small-24b-2507 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | mistral.voxtral-small-24b-2507 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-voxtral-small-24b-2507-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-voxtral-small-24b-2507-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-voxtral-small-24b-2507-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-voxtral-small-24b-2507-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.voxtral-small-24b-2507",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.voxtral-small-24b-2507",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.voxtral-small-24b-2507',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.voxtral-small-24b-2507',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Magistral Small 2509
<a name="model-card-mistral-ai-magistral-small-2509"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Magistral Small 2509
<a name="model-card-mistral-ai-magistral-small-2509-header"></a>

## Model Details
<a name="model-card-mistral-ai-magistral-small-2509-details"></a>

Magistral Small 2509 is Mistral AI's reasoning model that uses chain-of-thought to solve complex math, coding, and logic problems. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Sep 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 40K
+ **Reasoning:** Supported


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-magistral-small-2509-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-magistral-small-2509.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-magistral-small-2509.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-magistral-small-2509.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-magistral-small-2509.html)  | 

## Pricing
<a name="model-card-mistral-ai-magistral-small-2509-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-magistral-small-2509-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.magistral-small-2509 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | mistral.magistral-small-2509 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-magistral-small-2509-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-magistral-small-2509-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-magistral-small-2509-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-magistral-small-2509-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.magistral-small-2509",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.magistral-small-2509",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.magistral-small-2509',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.magistral-small-2509',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Voxtral Mini 3B 2507
<a name="model-card-mistral-ai-voxtral-mini-3b-2507"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Voxtral Mini 3B 2507
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-header"></a>

## Model Details
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-details"></a>

Voxtral Mini 3B is Mistral AI's compact speech-to-text model for real-time transcription and voice understanding on edge devices. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Jul 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 32K tokens


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-mini-3b-2507.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-mini-3b-2507.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-mini-3b-2507.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-voxtral-mini-3b-2507.html)  | 

## Pricing
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.voxtral-mini-3b-2507 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | mistral.voxtral-mini-3b-2507 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-voxtral-mini-3b-2507-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.voxtral-mini-3b-2507",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.voxtral-mini-3b-2507",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.voxtral-mini-3b-2507',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.voxtral-mini-3b-2507',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Devstral 2 123B
<a name="model-card-mistral-ai-devstral-2-123b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Devstral 2 123B
<a name="model-card-mistral-ai-devstral-2-123b-header"></a>

## Model Details
<a name="model-card-mistral-ai-devstral-2-123b-details"></a>

Devstral 2 123B is Mistral AI's 123-billion parameter coding model optimized for software engineering tasks including code generation, debugging, and refactoring. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Jun 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 32K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-mistral-ai-devstral-2-123b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-devstral-2-123b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-devstral-2-123b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-devstral-2-123b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-devstral-2-123b.html)  | 

## Pricing
<a name="model-card-mistral-ai-devstral-2-123b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-devstral-2-123b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.devstral-2-123b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | mistral.devstral-2-123b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-devstral-2-123b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-devstral-2-123b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-devstral-2-123b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-devstral-2-123b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.devstral-2-123b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.devstral-2-123b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.devstral-2-123b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.devstral-2-123b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Pixtral Large
<a name="model-card-mistral-ai-pixtral-large"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Pixtral Large
<a name="model-card-mistral-ai-pixtral-large-header"></a>

## Model Details
<a name="model-card-mistral-ai-pixtral-large-details"></a>

Pixtral Large is Mistral AI's 124-billion parameter multimodal model that processes text and images for visual reasoning and document understanding. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Nov 19, 2024
+ **Model EOL date:** No sooner than 4/8/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-mistral-ai-pixtral-large-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-pixtral-large-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.pixtral-large-2502-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.mistral.pixtral-large-2502-v1:0``eu.mistral.pixtral-large-2502-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-pixtral-large-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-pixtral-large-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.mistral.pixtral-large-2502-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.mistral.pixtral-large-2502-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-west-1 (Ireland), eu-west-3 (Paris) | 

## Quotas and Limits
<a name="model-card-mistral-ai-pixtral-large-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-pixtral-large-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.pixtral-large-2502-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.pixtral-large-2502-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Mistral Large
<a name="model-card-mistral-ai-mistral-large"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Mistral Large
<a name="model-card-mistral-ai-mistral-large-header"></a>

## Model Details
<a name="model-card-mistral-ai-mistral-large-details"></a>

Mistral Large is Mistral AI's flagship model with strong reasoning, multilingual support, and a 32K context window for complex enterprise tasks. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Feb 26, 2024
+ **Model EOL date:** No sooner than 4/3/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 32K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-mistral-ai-mistral-large-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-mistral-ai-mistral-large.html)  | 

## Pricing
<a name="model-card-mistral-ai-mistral-large-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-mistral-large-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.mistral-large-2402-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-mistral-large-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-mistral-large-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-mistral-large-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-mistral-large-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="mistral.mistral-large-3-675b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="mistral.mistral-large-3-675b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.mistral-large-2402-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.mistral-large-2402-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Mixtral 8x7B Instruct
<a name="model-card-mistral-ai-mixtral-8x7b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Mixtral 8x7B Instruct
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-header"></a>

## Model Details
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-details"></a>

Mixtral 8x7B Instruct is Mistral AI's sparse mixture-of-experts model with 8 experts and 7B parameters each, delivering strong performance at faster inference speeds. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Dec 10, 2023
+ **Model EOL date:** No sooner than 3/1/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 32K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.mixtral-8x7b-instruct-v0:1 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-mixtral-8x7b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.mixtral-8x7b-instruct-v0:1',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.mixtral-8x7b-instruct-v0:1',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Mistral 7B Instruct
<a name="model-card-mistral-ai-mistral-7b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/mistralai.png) Mistral AI — Mistral 7B Instruct
<a name="model-card-mistral-ai-mistral-7b-instruct-header"></a>

## Model Details
<a name="model-card-mistral-ai-mistral-7b-instruct-details"></a>

Mistral 7B Instruct is Mistral AI's 7-billion parameter instruction-tuned model with grouped-query attention and sliding window attention for efficient long-context inference. For more information about model development and performance, see the [model/service card](https://docs.mistral.ai/getting-started/models).
+ **Model launch date:** Sep 28, 2023
+ **Model EOL date:** No sooner than 3/1/2025
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 32K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Pricing
<a name="model-card-mistral-ai-mistral-7b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-mistral-ai-mistral-7b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | mistral.mistral-7b-instruct-v0:2 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-mistral-ai-mistral-7b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-mistral-ai-mistral-7b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ca-central-1 (Canada) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-3 (Paris) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-mistral-ai-mistral-7b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-mistral-ai-mistral-7b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='mistral.mistral-7b-instruct-v0:2',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='mistral.mistral-7b-instruct-v0:2',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Moonshot AI
<a name="model-cards-moonshot-ai"></a>

The following Moonshot AI models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Kimi K2.5](model-card-moonshot-ai-kimi-k2-5.md) | Kimi K2.5 is Moonshot AI's multimodal model with improved reasoning, coding, and multilingual capabilities. | 
| [Kimi K2 Thinking](model-card-moonshot-ai-kimi-k2-thinking.md) | Kimi K2 Thinking is Moonshot AI's reasoning model with chain-of-thought capabilities for complex problem solving in math, coding, and logic. | 

# Kimi K2.5
<a name="model-card-moonshot-ai-kimi-k2-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/kimik2.5.png) Moonshot AI — Kimi K2.5
<a name="model-card-moonshot-ai-kimi-k2-5-header"></a>

## Model Details
<a name="model-card-moonshot-ai-kimi-k2-5-details"></a>

Kimi K2.5 is Moonshot AI's multimodal model with improved reasoning, coding, and multilingual capabilities. For more information about model development and performance, see the [model/service card](https://platform.moonshot.ai/docs/guide/kimi-k2-5-quickstart#overview-of-kimi-k25-model).
+ **Model launch date:** Jan 27, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-moonshot-ai-kimi-k2-5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-5.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-5.html)  | 

## Pricing
<a name="model-card-moonshot-ai-kimi-k2-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-moonshot-ai-kimi-k2-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | moonshotai.kimi-k2.5 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | moonshotai.kimi-k2.5 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-moonshot-ai-kimi-k2-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-moonshot-ai-kimi-k2-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-moonshot-ai-kimi-k2-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-moonshot-ai-kimi-k2-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="moonshotai.kimi-k2.5",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="moonshotai.kimi-k2.5",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='moonshotai.kimi-k2.5',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='moonshotai.kimi-k2.5',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Kimi K2 Thinking
<a name="model-card-moonshot-ai-kimi-k2-thinking"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/kimik2.5.png) Moonshot AI — Kimi K2 Thinking
<a name="model-card-moonshot-ai-kimi-k2-thinking-header"></a>

## Model Details
<a name="model-card-moonshot-ai-kimi-k2-thinking-details"></a>

Kimi K2 Thinking is Moonshot AI's reasoning model with chain-of-thought capabilities for complex problem solving in math, coding, and logic. For more information about model development and performance, see the [model/service card](https://platform.moonshot.ai/docs/guide/kimi-k2-quickstart#overview-of-kimi-k2).
+ **Model launch date:** Nov 06, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K
+ **Reasoning:** Supported


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-moonshot-ai-kimi-k2-thinking-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-moonshot-ai-kimi-k2-thinking.html)  | 

## Pricing
<a name="model-card-moonshot-ai-kimi-k2-thinking-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-moonshot-ai-kimi-k2-thinking-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | moonshot.kimi-k2-thinking | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | moonshotai.kimi-k2-thinking | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-moonshot-ai-kimi-k2-thinking-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-moonshot-ai-kimi-k2-thinking-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-moonshot-ai-kimi-k2-thinking-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-moonshot-ai-kimi-k2-thinking-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="moonshotai.kimi-k2-thinking",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="moonshotai.kimi-k2-thinking",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='moonshot.kimi-k2-thinking',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='moonshot.kimi-k2-thinking',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# NVIDIA
<a name="model-cards-nvidia"></a>

The following NVIDIA models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [NVIDIA Nemotron Nano 9B v2](model-card-nvidia-nvidia-nemotron-nano-9b-v2.md) | Nemotron Nano 9B v2 is NVIDIA's 9-billion parameter model optimized for efficient text generation, reasoning, and coding on NVIDIA hardware. | 
| [NVIDIA Nemotron Nano 12B v2 VL BF16](model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.md) | Nemotron Nano 12B v2 VL is NVIDIA's 12-billion parameter vision-language model for multimodal tasks including image understanding and visual Q&A. | 
| [Nemotron Nano 3 30B](model-card-nvidia-nemotron-nano-3-30b.md) | Nemotron Nano 3 30B is NVIDIA's 30-billion parameter model with strong reasoning and coding performance, optimized for deployment on NVIDIA GPUs. | 
| [NVIDIA Nemotron 3 Super 120B](model-card-nvidia-nemotron-super-3-120b.md) | NVIDIA Nemotron 3 Super 120B is NVIDIA's open hybrid MoE model with 120B total parameters and 12B active, designed for complex multi-agent applications with a 1M token context window. | 

# NVIDIA Nemotron Nano 9B v2
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nvidia.png) NVIDIA — NVIDIA Nemotron Nano 9B v2
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-header"></a>

## Model Details
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-details"></a>

Nemotron Nano 9B v2 is NVIDIA's 9-billion parameter model optimized for efficient text generation, reasoning, and coding on NVIDIA hardware. For more information about model development and performance, see the [model/service card](https://build.nvidia.com/nvidia/nvidia-nemotron-nano-9b-v2/modelcard).
+ **Model launch date:** Aug 18, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-9b-v2.html)  | 

## Pricing
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | nvidia.nemotron-nano-9b-v2 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | nvidia.nemotron-nano-9b-v2 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-nvidia-nvidia-nemotron-nano-9b-v2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="nvidia.nemotron-nano-9b-v2",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="nvidia.nemotron-nano-9b-v2",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='nvidia.nemotron-nano-9b-v2',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='nvidia.nemotron-nano-9b-v2',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# NVIDIA Nemotron Nano 12B v2 VL BF16
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nvidia.png) NVIDIA — NVIDIA Nemotron Nano 12B v2 VL BF16
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-header"></a>

## Model Details
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-details"></a>

Nemotron Nano 12B v2 VL is NVIDIA's 12-billion parameter vision-language model for multimodal tasks including image understanding and visual Q&A. For more information about model development and performance, see the [model/service card](https://build.nvidia.com/nvidia/nemotron-nano-12b-v2-vl/modelcard).
+ **Model launch date:** Oct 28, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16.html)  | 

## Pricing
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | nvidia.nemotron-nano-12b-v2 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | nvidia.nemotron-nano-12b-v2 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-nvidia-nvidia-nemotron-nano-12b-v2-vl-bf16-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="nvidia.nemotron-nano-12b-v2",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="nvidia.nemotron-nano-12b-v2",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='nvidia.nemotron-nano-12b-v2',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='nvidia.nemotron-nano-12b-v2',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Nemotron Nano 3 30B
<a name="model-card-nvidia-nemotron-nano-3-30b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nvidia.png) NVIDIA — Nemotron Nano 3 30B
<a name="model-card-nvidia-nemotron-nano-3-30b-header"></a>

## Model Details
<a name="model-card-nvidia-nemotron-nano-3-30b-details"></a>

Nemotron Nano 3 30B is NVIDIA's 30-billion parameter model with strong reasoning and coding performance, optimized for deployment on NVIDIA GPUs. For more information about model development and performance, see the [model/service card](https://build.nvidia.com/nvidia/nemotron-3-nano-30b-a3b/modelcard).
+ **Model launch date:** Dec 15, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-nvidia-nemotron-nano-3-30b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-nano-3-30b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-nano-3-30b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-nano-3-30b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-nano-3-30b.html)  | 

## Pricing
<a name="model-card-nvidia-nemotron-nano-3-30b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-nvidia-nemotron-nano-3-30b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | nvidia.nemotron-nano-3-30b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | nvidia.nemotron-nano-3-30b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-nvidia-nemotron-nano-3-30b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-nvidia-nemotron-nano-3-30b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-nvidia-nemotron-nano-3-30b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-nvidia-nemotron-nano-3-30b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="nvidia.nemotron-nano-3-30b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="nvidia.nemotron-nano-3-30b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='nvidia.nemotron-nano-3-30b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='nvidia.nemotron-nano-3-30b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# NVIDIA Nemotron 3 Super 120B
<a name="model-card-nvidia-nemotron-super-3-120b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/nvidia.png) NVIDIA — NVIDIA Nemotron 3 Super 120B
<a name="model-card-nvidia-nemotron-super-3-120b-header"></a>

## Model Details
<a name="model-card-nvidia-nemotron-super-3-120b-details"></a>

NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. It delivers up to 7x higher throughput, providing fast, cost-efficient inference for agentic tasks. A long context window gives the model long-term memory, preventing AI agents from losing focus on long, multi-step tasks and ensuring high-accuracy results. Fully open with weights, datasets, and recipes, it allows easy customization and secure deployment. For more information about model development and performance, see the [model/service card](https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard).
+ **Model launch date:** Mar 11, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 256K tokens
+ **Max output tokens:** 32K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-nvidia-nemotron-super-3-120b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-super-3-120b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-super-3-120b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-super-3-120b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-nvidia-nemotron-super-3-120b.html)  | 

## Pricing
<a name="model-card-nvidia-nemotron-super-3-120b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-nvidia-nemotron-super-3-120b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | nvidia.nemotron-super-3-120b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | nvidia.nemotron-super-3-120b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-nvidia-nemotron-super-3-120b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-nvidia-nemotron-super-3-120b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-nvidia-nemotron-super-3-120b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-nvidia-nemotron-super-3-120b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="nvidia.nemotron-super-3-120b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="nvidia.nemotron-super-3-120b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='nvidia.nemotron-super-3-120b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='nvidia.nemotron-super-3-120b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# OpenAI
<a name="model-cards-openai"></a>

The following OpenAI models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [GPT OSS Safeguard 120B](model-card-openai-gpt-oss-safeguard-120b.md) | GPT OSS Safeguard 120B is OpenAI's 120-billion parameter open-source safety model for content moderation and guardrail enforcement in AI applications. | 
| [GPT OSS Safeguard 20B](model-card-openai-gpt-oss-safeguard-20b.md) | GPT OSS Safeguard 20B is OpenAI's compact 20-billion parameter open-source safety model for lightweight content moderation and guardrail tasks. | 
| [gpt-oss-120b](model-card-openai-gpt-oss-120b.md) | GPT OSS 120B is OpenAI's 120-billion parameter open-source general-purpose model for text generation, coding, and reasoning tasks. | 
| [gpt-oss-20b](model-card-openai-gpt-oss-20b.md) | GPT OSS 20B is OpenAI's 20-billion parameter open-source model for efficient text generation and coding at lower compute cost. | 

# GPT OSS Safeguard 120B
<a name="model-card-openai-gpt-oss-safeguard-120b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/openai.png) OpenAI — GPT OSS Safeguard 120B
<a name="model-card-openai-gpt-oss-safeguard-120b-header"></a>

## Model Details
<a name="model-card-openai-gpt-oss-safeguard-120b-details"></a>

GPT OSS Safeguard 120B is OpenAI's 120-billion parameter open-source safety model for content moderation and guardrail enforcement in AI applications. For more information about model development and performance, see the [model/service card](https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf).
+ **Model launch date:** Oct 29, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-openai-gpt-oss-safeguard-120b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-safeguard-120b.html)  |  —  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-safeguard-120b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-safeguard-120b.html)  | 

## Pricing
<a name="model-card-openai-gpt-oss-safeguard-120b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-openai-gpt-oss-safeguard-120b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | openai.gpt-oss-safeguard-120b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | openai.gpt-oss-safeguard-120b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-openai-gpt-oss-safeguard-120b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-openai-gpt-oss-safeguard-120b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-openai-gpt-oss-safeguard-120b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-openai-gpt-oss-safeguard-120b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="openai.gpt-oss-safeguard-120b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="openai.gpt-oss-safeguard-120b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='openai.gpt-oss-safeguard-120b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='openai.gpt-oss-safeguard-120b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# GPT OSS Safeguard 20B
<a name="model-card-openai-gpt-oss-safeguard-20b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/openai.png) OpenAI — GPT OSS Safeguard 20B
<a name="model-card-openai-gpt-oss-safeguard-20b-header"></a>

## Model Details
<a name="model-card-openai-gpt-oss-safeguard-20b-details"></a>

GPT OSS Safeguard 20B is OpenAI's compact 20-billion parameter open-source safety model for lightweight content moderation and guardrail tasks. For more information about model development and performance, see the [model/service card](https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf).
+ **Model launch date:** Oct 29, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-openai-gpt-oss-safeguard-20b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-safeguard-20b.html)  |  —  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-safeguard-20b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-safeguard-20b.html)  | 

## Pricing
<a name="model-card-openai-gpt-oss-safeguard-20b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-openai-gpt-oss-safeguard-20b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | openai.gpt-oss-safeguard-20b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | openai.gpt-oss-safeguard-20b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-openai-gpt-oss-safeguard-20b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-openai-gpt-oss-safeguard-20b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-openai-gpt-oss-safeguard-20b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-openai-gpt-oss-safeguard-20b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="openai.gpt-oss-safeguard-20b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="openai.gpt-oss-safeguard-20b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='openai.gpt-oss-safeguard-20b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='openai.gpt-oss-safeguard-20b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# gpt-oss-120b
<a name="model-card-openai-gpt-oss-120b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/openai.png) OpenAI — gpt-oss-120b
<a name="model-card-openai-gpt-oss-120b-header"></a>

## Model Details
<a name="model-card-openai-gpt-oss-120b-details"></a>

GPT OSS 120B is OpenAI's 120-billion parameter open-source general-purpose model for text generation, coding, and reasoning tasks. For more information about model development and performance, see the [model/service card](https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf).
+ **Model launch date:** Aug 05, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-openai-gpt-oss-120b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-120b.html)  |  —  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-120b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-120b.html)  | 

## Pricing
<a name="model-card-openai-gpt-oss-120b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-openai-gpt-oss-120b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | openai.gpt-oss-120b-1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | openai.gpt-oss-120b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-openai-gpt-oss-120b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-openai-gpt-oss-120b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-openai-gpt-oss-120b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-openai-gpt-oss-120b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="openai.gpt-oss-120b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="openai.gpt-oss-120b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='openai.gpt-oss-120b-1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='openai.gpt-oss-120b-1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# gpt-oss-20b
<a name="model-card-openai-gpt-oss-20b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/openai.png) OpenAI — gpt-oss-20b
<a name="model-card-openai-gpt-oss-20b-header"></a>

## Model Details
<a name="model-card-openai-gpt-oss-20b-details"></a>

GPT OSS 20B is OpenAI's 20-billion parameter open-source model for efficient text generation and coding at lower compute cost. For more information about model development and performance, see the [model/service card](https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf).
+ **Model launch date:** Aug 05, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-openai-gpt-oss-20b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-20b.html)  |  —  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-20b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-openai-gpt-oss-20b.html)  | 

## Pricing
<a name="model-card-openai-gpt-oss-20b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-openai-gpt-oss-20b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | openai.gpt-oss-20b-1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | openai.gpt-oss-20b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-openai-gpt-oss-20b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-openai-gpt-oss-20b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-openai-gpt-oss-20b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-openai-gpt-oss-20b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="openai.gpt-oss-20b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="openai.gpt-oss-20b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='openai.gpt-oss-20b-1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='openai.gpt-oss-20b-1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen
<a name="model-cards-qwen"></a>

The following Qwen models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Qwen3 Coder Next](model-card-qwen-qwen3-coder-next.md) | Qwen3 Coder Next is Qwen's coding model with improved code generation, debugging, and software engineering capabilities. | 
| [Qwen3 VL 235B A22B](model-card-qwen-qwen3-vl-235b-a22b.md) | Qwen3 VL 235B A22B is Qwen's vision-language mixture-of-experts model that processes text and images for visual reasoning and document understanding. | 
| [Qwen3 Next 80B A3B](model-card-qwen-qwen3-next-80b-a3b.md) | Qwen3 Next 80B A3B is Qwen's efficient mixture-of-experts model with 80B total and 3B active parameters for fast, cost-effective inference. | 
| [Qwen3-Coder-30B-A3B-Instruct](model-card-qwen-qwen3-coder-30b-a3b-instruct.md) | Qwen3 Coder 30B A3B is Qwen's compact coding mixture-of-experts model with 30B total and 3B active parameters for efficient code generation. | 
| [Qwen3 Coder 480B A35B Instruct](model-card-qwen-qwen3-coder-480b-a35b-instruct.md) | Qwen3 Coder 480B A35B is Qwen's largest coding-specialized mixture-of-experts model with 480B total and 35B active parameters for software engineering tasks. | 
| [Qwen3 32B](model-card-qwen-qwen3-32b.md) | Qwen3 32B is Qwen's 32-billion parameter dense model with hybrid thinking modes for both fast responses and deep reasoning. | 
| [Qwen3 235B A22B 2507](model-card-qwen-qwen3-235b-a22b-2507.md) | Qwen3 235B A22B is Qwen's 235-billion parameter mixture-of-experts model with 22 billion active parameters, supporting text and code generation with a 128K context window. | 

# Qwen3 Coder Next
<a name="model-card-qwen-qwen3-coder-next"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3 Coder Next
<a name="model-card-qwen-qwen3-coder-next-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-coder-next-details"></a>

Qwen3 Coder Next is Qwen's coding model with improved code generation, debugging, and software engineering capabilities. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3-coder).
+ **Model launch date:** Feb 04, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-coder-next-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-next.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-next.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-next.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-next.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-coder-next-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-coder-next-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-coder-next | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-coder-next | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-coder-next-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-coder-next-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-qwen-qwen3-coder-next-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-qwen-qwen3-coder-next-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-coder-next",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-coder-next",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-coder-next',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-coder-next',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen3 VL 235B A22B
<a name="model-card-qwen-qwen3-vl-235b-a22b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3 VL 235B A22B
<a name="model-card-qwen-qwen3-vl-235b-a22b-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-vl-235b-a22b-details"></a>

Qwen3 VL 235B A22B is Qwen's vision-language mixture-of-experts model that processes text and images for visual reasoning and document understanding. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3).
+ **Model launch date:** Sep 23, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-vl-235b-a22b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-vl-235b-a22b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-vl-235b-a22b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-vl-235b-a22b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-vl-235b-a22b.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-vl-235b-a22b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-vl-235b-a22b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-vl-235b-a22b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-vl-235b-a22b-instruct | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-vl-235b-a22b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-vl-235b-a22b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-qwen-qwen3-vl-235b-a22b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-qwen-qwen3-vl-235b-a22b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-vl-235b-a22b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-vl-235b-a22b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-vl-235b-a22b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-vl-235b-a22b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen3 Next 80B A3B
<a name="model-card-qwen-qwen3-next-80b-a3b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3 Next 80B A3B
<a name="model-card-qwen-qwen3-next-80b-a3b-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-next-80b-a3b-details"></a>

Qwen3 Next 80B A3B is Qwen's efficient mixture-of-experts model with 80B total and 3B active parameters for fast, cost-effective inference. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3).
+ **Model launch date:** Sep 11, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Reasoning:** Supported


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-next-80b-a3b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-next-80b-a3b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-next-80b-a3b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-next-80b-a3b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-next-80b-a3b.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-next-80b-a3b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-next-80b-a3b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-next-80b-a3b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-next-80b-a3b-instruct | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-next-80b-a3b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-next-80b-a3b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-qwen-qwen3-next-80b-a3b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-qwen-qwen3-next-80b-a3b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-next-80b-a3b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-next-80b-a3b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-next-80b-a3b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-next-80b-a3b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen3-Coder-30B-A3B-Instruct
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3-Coder-30B-A3B-Instruct
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-details"></a>

Qwen3 Coder 30B A3B is Qwen's compact coding mixture-of-experts model with 30B total and 3B active parameters for efficient code generation. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3-coder).
+ **Model launch date:** Jul 31, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-30b-a3b-instruct.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-30b-a3b-instruct.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-30b-a3b-instruct.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-30b-a3b-instruct.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-coder-30b-a3b-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-coder-30b-a3b-instruct | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-qwen-qwen3-coder-30b-a3b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-coder-30b-a3b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-coder-30b-a3b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-coder-30b-a3b-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-coder-30b-a3b-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen3 Coder 480B A35B Instruct
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3 Coder 480B A35B Instruct
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-details"></a>

Qwen3 Coder 480B A35B is Qwen's largest coding-specialized mixture-of-experts model with 480B total and 35B active parameters for software engineering tasks. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3-coder).
+ **Model launch date:** Jul 23, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 16K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-480b-a35b-instruct.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-coder-480b-a35b-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-coder-480b-a35b-instruct | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-qwen-qwen3-coder-480b-a35b-instruct-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-coder-480b-a35b-instruct",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-coder-480b-a35b-instruct",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-coder-480b-a35b-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-coder-480b-a35b-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen3 32B
<a name="model-card-qwen-qwen3-32b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3 32B
<a name="model-card-qwen-qwen3-32b-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-32b-details"></a>

Qwen3 32B is Qwen's 32-billion parameter dense model with hybrid thinking modes for both fast responses and deep reasoning. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3).
+ **Model launch date:** Apr 29, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Reasoning:** Supported


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-32b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-32b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-32b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-32b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-32b.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-32b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-32b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-32b-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-32b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-32b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-32b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-qwen-qwen3-32b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-qwen-qwen3-32b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-32b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-32b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-32b-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-32b-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Qwen3 235B A22B 2507
<a name="model-card-qwen-qwen3-235b-a22b-2507"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/qwen.png) Qwen — Qwen3 235B A22B 2507
<a name="model-card-qwen-qwen3-235b-a22b-2507-header"></a>

## Model Details
<a name="model-card-qwen-qwen3-235b-a22b-2507-details"></a>

Qwen3 235B A22B is Qwen's 235-billion parameter mixture-of-experts model with 22 billion active parameters, supporting text and code generation with a 128K context window. For more information about model development and performance, see the [model/service card](https://qwen.ai/blog?id=qwen3).
+ **Model launch date:** Apr 28, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Reasoning:** Supported


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-qwen-qwen3-235b-a22b-2507-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-235b-a22b-2507.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-235b-a22b-2507.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-235b-a22b-2507.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-235b-a22b-2507.html)  | 

## Pricing
<a name="model-card-qwen-qwen3-235b-a22b-2507-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-qwen-qwen3-235b-a22b-2507-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | qwen.qwen3-235b-a22b-2507-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | qwen.qwen3-235b-a22b-2507 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-qwen-qwen3-235b-a22b-2507-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-qwen-qwen3-235b-a22b-2507-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Sample Code
<a name="model-card-qwen-qwen3-235b-a22b-2507-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="qwen.qwen3-235b-a22b-2507",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="qwen.qwen3-235b-a22b-2507",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='qwen.qwen3-235b-a22b-2507-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='qwen.qwen3-235b-a22b-2507-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Stability AI
<a name="model-cards-stability-ai"></a>

The following Stability AI models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Stable Image Remove Background](model-card-stability-ai-stable-image-remove-background.md) | Stable Image Remove Background is Stability AI's model that accurately removes backgrounds from images, isolating the foreground subject. | 
| [Stable Image Search and Replace](model-card-stability-ai-stable-image-search-and-replace.md) | Stable Image Search and Replace is Stability AI's model that finds objects in images and replaces them with new content based on text prompts. | 
| [Stable Image Style Guide](model-card-stability-ai-stable-image-style-guide.md) | Stable Image Style Guide is Stability AI's model that generates images matching a reference style while following text prompt instructions. | 
| [Stable Image Style Transfer](model-card-stability-ai-stable-image-style-transfer.md) | Stable Image Style Transfer is Stability AI's model that applies the artistic style of a reference image to a target image. | 
| [Stable Image Fast Upscale](model-card-stability-ai-stable-image-fast-upscale.md) | Stable Image Fast Upscale is Stability AI's model that quickly increases image resolution with minimal processing time. | 
| [Stable Image Search and Recolor](model-card-stability-ai-stable-image-search-and-recolor.md) | Stable Image Search and Recolor is Stability AI's model that identifies objects in images and changes their colors based on text prompts. | 
| [Stable Image Conservative Upscale](model-card-stability-ai-stable-image-conservative-upscale.md) | Stable Image Conservative Upscale is Stability AI's image upscaling model that increases resolution while preserving the original image's details and style. | 
| [Stable Image Erase Object](model-card-stability-ai-stable-image-erase-object.md) | Stable Image Erase Object is Stability AI's model that removes unwanted objects from images and fills the area with contextually appropriate content. | 
| [Stable Image Control Structure](model-card-stability-ai-stable-image-control-structure.md) | Stable Image Control Structure is Stability AI's model that generates images guided by structural inputs like depth maps and edge detection. | 
| [Stable Image Creative Upscale](model-card-stability-ai-stable-image-creative-upscale.md) | Stable Image Creative Upscale is Stability AI's model that upscales images while adding creative detail and enhancing visual quality. | 
| [Stable Image Control Sketch](model-card-stability-ai-stable-image-control-sketch.md) | Stable Image Control Sketch is Stability AI's model that generates images guided by sketch inputs for controlled image creation. | 
| [Stable Image Outpaint](model-card-stability-ai-stable-image-outpaint.md) | Stable Image Outpaint is Stability AI's model that extends images beyond their original boundaries with contextually coherent content. | 
| [Stable Image Inpaint](model-card-stability-ai-stable-image-inpaint.md) | Stable Image Inpaint is Stability AI's model that fills in masked regions of images with contextually appropriate content based on text prompts. | 

# Stable Image Remove Background
<a name="model-card-stability-ai-stable-image-remove-background"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Remove Background
<a name="model-card-stability-ai-stable-image-remove-background-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-remove-background-details"></a>

Stable Image Remove Background is Stability AI's model that accurately removes backgrounds from images, isolating the foreground subject. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Nov 7, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-remove-background-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-remove-background.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-remove-background.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-remove-background-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-remove-background-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-remove-background-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-remove-background-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-remove-background-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-remove-background-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-remove-background-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-remove-background-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-remove-background-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-remove-background-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Search and Replace
<a name="model-card-stability-ai-stable-image-search-and-replace"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Search and Replace
<a name="model-card-stability-ai-stable-image-search-and-replace-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-search-and-replace-details"></a>

Stable Image Search and Replace is Stability AI's model that finds objects in images and replaces them with new content based on text prompts. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Oct 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-search-and-replace-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-search-and-replace.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-search-and-replace.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-search-and-replace-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-search-and-replace-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-search-replace-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-search-replace-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-search-and-replace-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-search-and-replace-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-search-replace-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-search-and-replace-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-search-and-replace-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-search-replace-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Style Guide
<a name="model-card-stability-ai-stable-image-style-guide"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Style Guide
<a name="model-card-stability-ai-stable-image-style-guide-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-style-guide-details"></a>

Stable Image Style Guide is Stability AI's model that generates images matching a reference style while following text prompt instructions. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Oct 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-style-guide-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-style-guide.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-style-guide.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-style-guide-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-style-guide-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-style-guide-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-style-guide-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-style-guide-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-style-guide-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-style-guide-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-style-guide-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-style-guide-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-style-guide-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Style Transfer
<a name="model-card-stability-ai-stable-image-style-transfer"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Style Transfer
<a name="model-card-stability-ai-stable-image-style-transfer-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-style-transfer-details"></a>

Stable Image Style Transfer is Stability AI's model that applies the artistic style of a reference image to a target image. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Oct 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-style-transfer-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-style-transfer.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-style-transfer.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-style-transfer-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-style-transfer-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-style-transfer-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-style-transfer-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-style-transfer-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-style-transfer-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-style-transfer-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-style-transfer-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-style-transfer-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-style-transfer-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Fast Upscale
<a name="model-card-stability-ai-stable-image-fast-upscale"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Fast Upscale
<a name="model-card-stability-ai-stable-image-fast-upscale-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-fast-upscale-details"></a>

Stable Image Fast Upscale is Stability AI's model that quickly increases image resolution with minimal processing time. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Sep 10, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-fast-upscale-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-fast-upscale.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-fast-upscale.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-fast-upscale-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-fast-upscale-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-fast-upscale-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-fast-upscale-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-fast-upscale-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-fast-upscale-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-fast-upscale-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-fast-upscale-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-fast-upscale-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-fast-upscale-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Search and Recolor
<a name="model-card-stability-ai-stable-image-search-and-recolor"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Search and Recolor
<a name="model-card-stability-ai-stable-image-search-and-recolor-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-search-and-recolor-details"></a>

Stable Image Search and Recolor is Stability AI's model that identifies objects in images and changes their colors based on text prompts. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Aug 19, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-search-and-recolor-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-search-and-recolor.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-search-and-recolor.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-search-and-recolor-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-search-and-recolor-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-search-recolor-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-search-recolor-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-search-and-recolor-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-search-and-recolor-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-search-recolor-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-search-and-recolor-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-search-and-recolor-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-search-recolor-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Conservative Upscale
<a name="model-card-stability-ai-stable-image-conservative-upscale"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Conservative Upscale
<a name="model-card-stability-ai-stable-image-conservative-upscale-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-conservative-upscale-details"></a>

Stable Image Conservative Upscale is Stability AI's image upscaling model that increases resolution while preserving the original image's details and style. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** May 20, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-conservative-upscale-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-conservative-upscale.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-conservative-upscale.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-conservative-upscale-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-conservative-upscale-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-conservative-upscale-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-conservative-upscale-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-conservative-upscale-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-conservative-upscale-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-conservative-upscale-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-conservative-upscale-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-conservative-upscale-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-conservative-upscale-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Erase Object
<a name="model-card-stability-ai-stable-image-erase-object"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Erase Object
<a name="model-card-stability-ai-stable-image-erase-object-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-erase-object-details"></a>

Stable Image Erase Object is Stability AI's model that removes unwanted objects from images and fills the area with contextually appropriate content. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** May 20, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-erase-object-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-erase-object.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-erase-object.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-erase-object-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-erase-object-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-erase-object-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-erase-object-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-erase-object-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-erase-object-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-erase-object-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-erase-object-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-erase-object-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-erase-object-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Control Structure
<a name="model-card-stability-ai-stable-image-control-structure"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Control Structure
<a name="model-card-stability-ai-stable-image-control-structure-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-control-structure-details"></a>

Stable Image Control Structure is Stability AI's model that generates images guided by structural inputs like depth maps and edge detection. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Apr 30, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-control-structure-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-control-structure.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-control-structure.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-control-structure-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-control-structure-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-control-structure-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-control-structure-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-control-structure-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-control-structure-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-control-structure-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-control-structure-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-control-structure-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-control-structure-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Creative Upscale
<a name="model-card-stability-ai-stable-image-creative-upscale"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Creative Upscale
<a name="model-card-stability-ai-stable-image-creative-upscale-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-creative-upscale-details"></a>

Stable Image Creative Upscale is Stability AI's model that upscales images while adding creative detail and enhancing visual quality. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Mar 21, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-creative-upscale-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-creative-upscale.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-creative-upscale.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-creative-upscale-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-creative-upscale-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-creative-upscale-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-creative-upscale-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-creative-upscale-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-creative-upscale-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-creative-upscale-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-creative-upscale-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-creative-upscale-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-creative-upscale-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Control Sketch
<a name="model-card-stability-ai-stable-image-control-sketch"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Control Sketch
<a name="model-card-stability-ai-stable-image-control-sketch-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-control-sketch-details"></a>

Stable Image Control Sketch is Stability AI's model that generates images guided by sketch inputs for controlled image creation. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Jul 13, 2023
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-control-sketch-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-control-sketch.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-control-sketch.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-control-sketch-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-control-sketch-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-control-sketch-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-control-sketch-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-control-sketch-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-control-sketch-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-control-sketch-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-control-sketch-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-control-sketch-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-control-sketch-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Outpaint
<a name="model-card-stability-ai-stable-image-outpaint"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Outpaint
<a name="model-card-stability-ai-stable-image-outpaint-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-outpaint-details"></a>

Stable Image Outpaint is Stability AI's model that extends images beyond their original boundaries with contextually coherent content. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Nov 24, 2022
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-outpaint-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-outpaint.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-outpaint.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-outpaint-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-outpaint-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-outpaint-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-outpaint-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-outpaint-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-outpaint-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-outpaint-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-outpaint-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-outpaint-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-outpaint-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Stable Image Inpaint
<a name="model-card-stability-ai-stable-image-inpaint"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/stability.png) Stability AI — Stable Image Inpaint
<a name="model-card-stability-ai-stable-image-inpaint-header"></a>

## Model Details
<a name="model-card-stability-ai-stable-image-inpaint-details"></a>

Stable Image Inpaint is Stability AI's model that fills in masked regions of images with contextually appropriate content based on text prompts. For more information about model development and performance, see the [model/service card](https://stability.ai/stable-image).
+ **Model launch date:** Nov 2022
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-stability-ai-stable-image-inpaint-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-inpaint.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-stability-ai-stable-image-inpaint.html)  | 

## Pricing
<a name="model-card-stability-ai-stable-image-inpaint-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-stability-ai-stable-image-inpaint-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | stability.stable-image-inpaint-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.stability.stable-image-inpaint-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-stability-ai-stable-image-inpaint-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-stability-ai-stable-image-inpaint-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.stability.stable-image-inpaint-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Quotas and Limits
<a name="model-card-stability-ai-stable-image-inpaint-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-stability-ai-stable-image-inpaint-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='stability.stable-image-inpaint-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# TwelveLabs
<a name="model-cards-twelvelabs"></a>

The following TwelveLabs models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Marengo Embed 3.0](model-card-twelvelabs-marengo-embed-3-0.md) | Marengo Embed 3.0 is TwelveLabs' video embedding model that generates vector representations of video content for search and retrieval. | 
| [Pegasus v1.2](model-card-twelvelabs-pegasus-v1-2.md) | Pegasus v1.2 is TwelveLabs' video-to-text generation model that produces detailed descriptions, summaries, and answers about video content. | 
| [Marengo Embed v2.7](model-card-twelvelabs-marengo-embed-v2-7.md) | Marengo Embed v2.7 is TwelveLabs' video embedding model for multimodal video understanding, search, and classification. | 

# Marengo Embed 3.0
<a name="model-card-twelvelabs-marengo-embed-3-0"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/twelvelabs.png) TwelveLabs — Marengo Embed 3.0
<a name="model-card-twelvelabs-marengo-embed-3-0-header"></a>

## Model Details
<a name="model-card-twelvelabs-marengo-embed-3-0-details"></a>

Marengo Embed 3.0 is TwelveLabs' video embedding model that generates vector representations of video content for search and retrieval. For more information about model development and performance, see the [model/service card](https://docs.twelvelabs.io/docs/concepts/models/marengo).
+ **Model launch date:** Oct 29, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Marketplace product ID:** `prod-ibqpnh3htjd4s`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) StartAsyncInvoke |  | 

## Capabilities and Features
<a name="model-card-twelvelabs-marengo-embed-3-0-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-marengo-embed-3-0.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-marengo-embed-3-0.html)  | 

## Pricing
<a name="model-card-twelvelabs-marengo-embed-3-0-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-twelvelabs-marengo-embed-3-0-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | twelvelabs.marengo-embed-3-0-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.twelvelabs.marengo-embed-3-0-v1:0``eu.twelvelabs.marengo-embed-3-0-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-twelvelabs-marengo-embed-3-0-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-twelvelabs-marengo-embed-3-0-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-2 (Seoul) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.twelvelabs.marengo-embed-3-0-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.twelvelabs.marengo-embed-3-0-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

## Quotas and Limits
<a name="model-card-twelvelabs-marengo-embed-3-0-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-twelvelabs-marengo-embed-3-0-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:**

```
pip install boto3
```

**Step 4 - Set environment variables:**

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** This model uses `StartAsyncInvoke`. Save the file as `bedrock-first-request.py`

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.start_async_invoke(
    modelId='twelvelabs.marengo-embed-3-0-v1:0',
    modelInput={},
    outputDataConfig={'s3OutputDataConfig': {'s3Uri': 's3://your-bucket/output/'}}
)
print(response)
```

# Pegasus v1.2
<a name="model-card-twelvelabs-pegasus-v1-2"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/twelvelabs.png) TwelveLabs — Pegasus v1.2
<a name="model-card-twelvelabs-pegasus-v1-2-header"></a>

## Model Details
<a name="model-card-twelvelabs-pegasus-v1-2-details"></a>

Pegasus v1.2 is TwelveLabs' video-to-text generation model that produces detailed descriptions, summaries, and answers about video content. For more information about model development and performance, see the [model/service card](https://docs.twelvelabs.io/docs/concepts/models/pegasus).
+ **Model launch date:** Feb 11, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Marketplace product ID:** `prod-635pcy5x5pc2a`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-twelvelabs-pegasus-v1-2-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.html)  | 

## Pricing
<a name="model-card-twelvelabs-pegasus-v1-2-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-twelvelabs-pegasus-v1-2-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | twelvelabs.pegasus-1-2-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.twelvelabs.pegasus-1-2-v1:0``eu.twelvelabs.pegasus-1-2-v1:0` | global.twelvelabs.pegasus-1-2-v1:0 | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-twelvelabs-pegasus-v1-2-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-twelvelabs-pegasus-v1-2-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-central-1 (Canada) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ca-west-1 (Calgary) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-1 (Frankfurt) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-central-2 (Zurich) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-north-1 (Stockholm) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-1 (Milan) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-south-2 (Spain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-2 (London) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| eu-west-3 (Paris) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-east-2 (Taipei) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-1 (Tokyo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-2 (Seoul) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-northeast-3 (Osaka) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-1 (Mumbai) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-south-2 (Hyderabad) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-1 (Singapore) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-2 (Sydney) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-3 (Jakarta) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-4 (Melbourne) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-5 (Malaysia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| ap-southeast-7 (Thailand) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| il-central-1 (Tel Aviv) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-central-1 (UAE) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| me-south-1 (Bahrain) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| af-south-1 (Cape Town) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| sa-east-1 (São Paulo) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 
| mx-central-1 (Mexico) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.twelvelabs.pegasus-1-2-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.twelvelabs.pegasus-1-2-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-central-1 (Frankfurt) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-central-2 (Zurich) | eu-central-1 (Frankfurt), eu-central-2 (Zurich), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-north-1 (Stockholm) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-1 (Milan) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-south-2 (Spain) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 
| eu-west-2 (London) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris) | 
| eu-west-3 (Paris) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

***Global inference details***


| **Global Inference ID** | **Americas** | **EMEA** | **Asia Pacific** | 
| --- | --- | --- | --- | 
| global.twelvelabs.pegasus-1-2-v1:0 |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-twelvelabs-pegasus-v1-2.html)  | 

## Quotas and Limits
<a name="model-card-twelvelabs-pegasus-v1-2-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-twelvelabs-pegasus-v1-2-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='twelvelabs.pegasus-1-2-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------

# Marengo Embed v2.7
<a name="model-card-twelvelabs-marengo-embed-v2-7"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/twelvelabs.png) TwelveLabs — Marengo Embed v2.7
<a name="model-card-twelvelabs-marengo-embed-v2-7-header"></a>

## Model Details
<a name="model-card-twelvelabs-marengo-embed-v2-7-details"></a>

Marengo Embed v2.7 is TwelveLabs' video embedding model for multimodal video understanding, search, and classification. For more information about model development and performance, see the [model/service card](https://docs.twelvelabs.io/docs/concepts/models/marengo).
+ **Model launch date:** Dec 4, 2024
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Marketplace product ID:** `prod-o6xchhpirymvs`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Text | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Converse |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) StartAsyncInvoke |  | 

## Pricing
<a name="model-card-twelvelabs-marengo-embed-v2-7-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-twelvelabs-marengo-embed-v2-7-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | twelvelabs.marengo-embed-2-7-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | `us.twelvelabs.marengo-embed-2-7-v1:0``eu.twelvelabs.marengo-embed-2-7-v1:0` | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-twelvelabs-marengo-embed-v2-7-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-twelvelabs-marengo-embed-v2-7-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.twelvelabs.marengo-embed-2-7-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 

**Geo: EU**

Geo Inference ID: `eu.twelvelabs.marengo-embed-2-7-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| eu-west-1 (Ireland) | eu-central-1 (Frankfurt), eu-north-1 (Stockholm), eu-south-1 (Milan), eu-south-2 (Spain), eu-west-1 (Ireland), eu-west-3 (Paris) | 

## Quotas and Limits
<a name="model-card-twelvelabs-marengo-embed-v2-7-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-twelvelabs-marengo-embed-v2-7-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:**

```
pip install boto3
```

**Step 4 - Set environment variables:**

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** This model uses `StartAsyncInvoke`. Save the file as `bedrock-first-request.py`

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.start_async_invoke(
    modelId='twelvelabs.marengo-embed-2-7-v1:0',
    modelInput={},
    outputDataConfig={'s3OutputDataConfig': {'s3Uri': 's3://your-bucket/output/'}}
)
print(response)
```

# Writer
<a name="model-cards-writer"></a>

The following Writer models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [Palmyra Vision 7B](model-card-writer-palmyra-vision-7b.md) | Palmyra Vision 7B is Writer's advanced multimodal language model for visual analysis, excelling at image understanding, handwritten text extraction, and chart interpretation. | 
| [Palmyra X5](model-card-writer-palmyra-x5.md) | Palmyra X5 is Writer's enterprise model with improved reasoning, coding, and agentic capabilities for complex business workflows. | 
| [Palmyra X4](model-card-writer-palmyra-x4.md) | Palmyra X4 is Writer's enterprise LLM optimized for business writing, content generation, and knowledge work with strong instruction following. | 

# Palmyra Vision 7B
<a name="model-card-writer-palmyra-vision-7b"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/writer.png) Writer — Palmyra Vision 7B
<a name="model-card-writer-palmyra-vision-7b-header"></a>

## Model Details
<a name="model-card-writer-palmyra-vision-7b-details"></a>

Palmyra Vision 7B is Writer's advanced multimodal language model, designed to interpret and generate text from images and video, providing robust visual analysis capabilities for enterprise needs. It excels at extracting handwritten text, interpreting complex charts and graphs, image-based compliance checks, and product description generation. For more information about model development and performance, see the [model/service card](https://writer.com/llms/palmyra-vision/).
+ **Model launch date:** Mar 26, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 4K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-writer-palmyra-vision-7b-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-vision-7b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-vision-7b.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-vision-7b.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-vision-7b.html)  | 

## Pricing
<a name="model-card-writer-palmyra-vision-7b-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-writer-palmyra-vision-7b-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | writer.palmyra-vision-7b | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | writer.palmyra-vision-7b | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-writer-palmyra-vision-7b-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-writer-palmyra-vision-7b-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-writer-palmyra-vision-7b-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-writer-palmyra-vision-7b-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="writer.palmyra-vision-7b",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="writer.palmyra-vision-7b",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='writer.palmyra-vision-7b',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='writer.palmyra-vision-7b',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Palmyra X5
<a name="model-card-writer-palmyra-x5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/writer.png) Writer — Palmyra X5
<a name="model-card-writer-palmyra-x5-header"></a>

## Model Details
<a name="model-card-writer-palmyra-x5-details"></a>

Palmyra X5 is Writer's enterprise model with improved reasoning, coding, and agentic capabilities for complex business workflows. For more information about model development and performance, see the [model/service card](https://writer.com/llms/palmyra-x5/).
+ **Model launch date:** Jan 21, 2026
+ **Model EOL date:** No sooner than 4/28/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Marketplace product ID:** `prod-23enyy63orhuk`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-writer-palmyra-x5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x5.html)  | 

## Pricing
<a name="model-card-writer-palmyra-x5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-writer-palmyra-x5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | writer.palmyra-x5-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.writer.palmyra-x5-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-writer-palmyra-x5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-writer-palmyra-x5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.writer.palmyra-x5-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Sample Code
<a name="model-card-writer-palmyra-x5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='writer.palmyra-x5-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='writer.palmyra-x5-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Palmyra X4
<a name="model-card-writer-palmyra-x4"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/writer.png) Writer — Palmyra X4
<a name="model-card-writer-palmyra-x4-header"></a>

## Model Details
<a name="model-card-writer-palmyra-x4-details"></a>

Palmyra X4 is Writer's enterprise LLM optimized for business writing, content generation, and knowledge work with strong instruction following. For more information about model development and performance, see the [model/service card](https://writer.com/llms/palmyra-x4/).
+ **Model launch date:** Sep 2024
+ **Model EOL date:** No sooner than 4/28/2026
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 8K
+ **Marketplace product ID:** `prod-azehe4da4pzsy`


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Chat Completions | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

## Capabilities and Features
<a name="model-card-writer-palmyra-x4-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x4.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-writer-palmyra-x4.html)  | 

## Pricing
<a name="model-card-writer-palmyra-x4-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-writer-palmyra-x4-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | writer.palmyra-x4-v1:0 | https://bedrock-runtime.\$1region\$1.amazonaws.com | us.writer.palmyra-x4-v1:0 | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-writer-palmyra-x4-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-writer-palmyra-x4-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-1 (N. California) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

***Geo inference details***

**Geo: US**

Geo Inference ID: `us.writer.palmyra-x4-v1:0`


| **Source Region** | **Destination Regions** | 
| --- | --- | 
| us-east-1 (N. Virginia) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-east-2 (Ohio) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 
| us-west-1 (N. California) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-1 (N. California), us-west-2 (Oregon) | 
| us-west-2 (Oregon) | us-east-1 (N. Virginia), us-east-2 (Ohio), us-west-2 (Oregon) | 

## Sample Code
<a name="model-card-writer-palmyra-x4-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

```
pip install boto3
```

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='writer.palmyra-x4-v1:0',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='writer.palmyra-x4-v1:0',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# Z.AI
<a name="model-cards-zai"></a>

The following Z.AI models are available in Amazon Bedrock:


| **Model** | **Description** | 
| --- | --- | 
| [GLM 4.7](model-card-zai-glm-4-7.md) | GLM 4.7 is Z.AI's large language model with strong multilingual capabilities and solid performance on reasoning, coding, and knowledge benchmarks. | 
| [GLM 4.7 Flash](model-card-zai-glm-4-7-flash.md) | GLM 4.7 Flash is Z.AI's lightweight model optimized for fast inference and low-latency tasks while maintaining strong general capabilities. | 
| [GLM 5](model-card-zai-glm-5.md) | GLM 5 is Z.AI's frontier-class general-purpose model with 744B parameters and 40B active, optimized for complex systems engineering and long-horizon agentic tasks. | 

# GLM 4.7
<a name="model-card-zai-glm-4-7"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/zai.png) Z.AI — GLM 4.7
<a name="model-card-zai-glm-4-7-header"></a>

## Model Details
<a name="model-card-zai-glm-4-7-details"></a>

GLM 4.7 is Z.AI's large language model with strong multilingual capabilities and solid performance on reasoning, coding, and knowledge benchmarks. For more information about model development and performance, see the [model/service card](https://z.ai/blog/glm-4.7).
+ **Model launch date:** Dec 22, 2025
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-zai-glm-4-7-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7.html)  | 

## Pricing
<a name="model-card-zai-glm-4-7-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-zai-glm-4-7-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | zai.glm-4.7 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | zai.glm-4.7 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-zai-glm-4-7-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-zai-glm-4-7-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-zai-glm-4-7-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-zai-glm-4-7-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="zai.glm-4.7",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="zai.glm-4.7",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='zai.glm-4.7',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='zai.glm-4.7',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# GLM 4.7 Flash
<a name="model-card-zai-glm-4-7-flash"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/zai.png) Z.AI — GLM 4.7 Flash
<a name="model-card-zai-glm-4-7-flash-header"></a>

## Model Details
<a name="model-card-zai-glm-4-7-flash-details"></a>

GLM 4.7 Flash is Z.AI's lightweight model optimized for fast inference and low-latency tasks while maintaining strong general capabilities. For more information about model development and performance, see the [model/service card](https://z.ai/blog/glm-4.7).
+ **Model launch date:** Jan 19, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 128K tokens
+ **Max output tokens:** 4K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-zai-glm-4-7-flash-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7-flash.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7-flash.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7-flash.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-4-7-flash.html)  | 

## Pricing
<a name="model-card-zai-glm-4-7-flash-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-zai-glm-4-7-flash-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | zai.glm-4.7-flash | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | zai.glm-4.7-flash | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-zai-glm-4-7-flash-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-zai-glm-4-7-flash-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-central-1 (Frankfurt) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-south-1 (Milan) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-1 (Ireland) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-zai-glm-4-7-flash-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-zai-glm-4-7-flash-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="zai.glm-4.7-flash",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="zai.glm-4.7-flash",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='zai.glm-4.7-flash',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='zai.glm-4.7-flash',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------

# GLM 5
<a name="model-card-zai-glm-5"></a>

## ![\[alt text not found\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/models/zai.png) Z.AI — GLM 5
<a name="model-card-zai-glm-5-header"></a>

## Model Details
<a name="model-card-zai-glm-5-details"></a>

GLM 5 is a frontier-class, general-purpose large language model optimized for complex systems engineering and long-horizon agentic tasks. It builds on the GLM 4.5 agent-centric lineage and is designed to support multi-step reasoning, math (including AIME-style benchmarks), advanced coding, and tool-augmented workflows, with long context support suitable for sophisticated agents and enterprise applications. For more information about model development and performance, see the [model/service card](https://docs.z.ai/guides/llm/glm-5-turbo).
+ **Model launch date:** Feb 11, 2026
+ **Model EOL date:** N/A
+ **End User License Agreements and Terms of Use:** [View](https://aws.amazon.com/legal/bedrock/third-party-models/)
+ **Model lifecycle:** Active
+ **Context window:** 200K tokens
+ **Max output tokens:** 128K


| **Input Modalities** | **Output Modalities** | **[APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html)** | **[Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html)** | 
| --- | --- | --- | --- | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Audio | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Embedding | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Responses | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-runtime | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Image | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Chat Completions | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) bedrock-mantle | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Speech | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Invoke |  | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Text | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) Converse |  | 
| ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) Video |  |  | 

**Note**  
Whenever possible, we recommend you use the `bedrock-mantle` endpoint.

## Capabilities and Features
<a name="model-card-zai-glm-5-capabilities"></a>

***Bedrock Features***

**Features supported using `bedrock-mantle` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-5.html)  | 

**Features supported using `bedrock-runtime` endpoint**


| **Supported** | **Not Supported** | 
| --- | --- | 
|  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-5.html)  |  [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/bedrock/latest/userguide/model-card-zai-glm-5.html)  | 

## Pricing
<a name="model-card-zai-glm-5-pricing"></a>

For pricing, please refer to the [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.

## Programmatic Access
<a name="model-card-zai-glm-5-programmatic-access"></a>

Use the following model IDs and endpoint URLs to access this model programmatically. For more information about the available APIs and endpoints, see [APIs supported](https://docs.aws.amazon.com/bedrock/latest/userguide/apis.html) and [Endpoints supported](https://docs.aws.amazon.com/bedrock/latest/userguide/endpoints.html).


| **Endpoint** | **Model ID** | **In-Region endpoint URL** | **Geo inference ID** | **Global inference ID** | 
| --- | --- | --- | --- | --- | 
| bedrock-runtime | zai.glm-5 | https://bedrock-runtime.\$1region\$1.amazonaws.com | Not supported | Not supported | 
| bedrock-mantle | zai.glm-5 | https://bedrock-mantle.\$1region\$1.api.aws/v1 | Not supported | Not supported | 

*For example, if region is us-east-1 (N. Virginia), then the bedrock-runtime endpoint URL will be "https://bedrock-runtime.us-east-1.amazonaws.com" and for bedrock-mantle will be "https://bedrock-mantle.us-east-1.api.aws/v1".*

## Service Tiers
<a name="model-card-zai-glm-5-tiers"></a>

Amazon Bedrock offers multiple service tiers to match your workload requirements. **Standard** provides pay-per-token access with no commitment. **Priority** offers higher throughput with a time-based commitment. **Flex** provides lower-cost access for flexible, non-time-sensitive workloads. **Reserved** provides dedicated throughput with a term commitment for predictable workloads. For more information, see [service tiers](https://docs.aws.amazon.com/bedrock/latest/userguide/service-tiers-inference.html).


| **Standard** | **Priority** | **Flex** | **Reserved** | 
| --- | --- | --- | --- | 
| ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Regional Availability
<a name="model-card-zai-glm-5-regional-availability"></a>

***Regional availability at a glance***

Bedrock offers three inference options: **In-Region** keeps requests within a single Region for strict compliance, **Geo Cross-Region** routes across Regions within a geography (US, EU, etc.) for higher throughput while respecting data residency, and **Global Cross-Region** routes anywhere worldwide for maximum throughput when there are no residency constraints. Refer to the [Regional availability](models-region-compatibility.md) page for more details.


| **Region** | **In-Region** | **Geo** | **Global** | 
| --- | --- | --- | --- | 
| us-east-1 (N. Virginia) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-east-2 (Ohio) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| us-west-2 (Oregon) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-north-1 (Stockholm) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| eu-west-2 (London) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-northeast-1 (Tokyo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-south-1 (Mumbai) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-2 (Sydney) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| ap-southeast-3 (Jakarta) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 
| sa-east-1 (São Paulo) | ![\[Yes\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-yes.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | ![\[No\]](http://docs.aws.amazon.com/bedrock/latest/userguide/images/icons/icon-no.png) | 

## Quotas and Limits
<a name="model-card-zai-glm-5-quotas"></a>

Your AWS account has default quotas to maintain the performance of the service and to ensure appropriate usage of Amazon Bedrock. The default quotas assigned to an account might be updated depending on regional factors, payment history, fraudulent usage, and/or approval of a quota [increase request](https://docs.aws.amazon.com/bedrock/latest/userguide/quotas-increase.html). For more details, please refer to [Quotas for Amazon Bedrock](quotas.md) documentation and see the [limits](https://docs.aws.amazon.com/general/latest/gr/bedrock.html#limits_bedrock) for the model.

## Sample Code
<a name="model-card-zai-glm-5-sample-code"></a>

**Step 1 - AWS Account:** If you have an AWS account already, skip this step. If you are new to AWS, sign up for an [AWS account](https://portal.aws.amazon.com/billing/signup).

**Step 2 - API key:** Go to the [Amazon Bedrock console](https://console.aws.amazon.com/bedrock/home#/api-keys/long-term/create) and generate a long-term API key.

**Step 3 - Get the SDK:** To use this getting started guide, you must have Python already installed. Then install the relevant software depending on the APIs you are using.

------
#### [ Responses/Chat Completions API ]

```
pip install boto3 openai
```

------
#### [ Invoke/Converse API ]

```
pip install boto3
```

------

**Step 4 - Set environment variables:** Configure your environment to use the API key for authentication.

------
#### [ Responses/Chat Completions API ]

```
OPENAI_API_KEY="<provide your Bedrock API key>"
OPENAI_BASE_URL="https://bedrock-mantle.<your-region>.api.aws/v1"
```

------
#### [ Invoke/Converse API ]

```
AWS_BEARER_TOKEN_BEDROCK="<provide your Bedrock API key>"
```

------

**Step 5 - Run your first inference request:** Save the file as `bedrock-first-request.py`

------
#### [ Responses API ]

```
from openai import OpenAI

client = OpenAI()

response = client.responses.create(
    model="zai.glm-5",
    input="Can you explain the features of Amazon Bedrock?"
    )
print(response)
```

------
#### [ Chat Completions API ]

```
from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="zai.glm-5",
    messages=[{"role": "user", "content": "Can you explain the features of Amazon Bedrock?"}]
    )
print(response)
```

------
#### [ Invoke API ]

```
import json
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.invoke_model(
    modelId='zai.glm-5',
    body=json.dumps({
            'messages': [{ 'role': 'user', 'content': 'Can you explain the features of Amazon Bedrock?'}],
            'max_tokens': 1024
    })
 )
 print(json.loads(response['body'].read()))
```

------
#### [ Converse API ]

```
import boto3

client = boto3.client('bedrock-runtime', region_name='us-east-1')
response = client.converse(
    modelId='zai.glm-5',
    messages=[
        {
            'role': 'user',
            'content': [{'text': 'Can you explain the features of Amazon Bedrock?'}]
        }
    ]
)
print(response)
```

------