Welcome
This guide provides detailed information on the Amazon Transcribe API, including operations, request and response syntax, data types, and error codes. For more information about Amazon Transcribe, refer to the Amazon Transcribe Developer Guide.
The Amazon Transcribe API reference is separated into the following sections:
If you're transcribing a media file located in an Amazon S3 bucket, you're performing a batch transcription and must use the operations and data types listed in the Amazon Transcribe Service section.
If you're transcribing a real-time media stream, you're performing a streaming transcription and must use the operations and data types listed in the Amazon Transcribe Streaming Service section.
Amazon Transcribe supports the following SDKs:
Batch transcriptions |
Streaming transcriptions |
|---|---|
The CLI is not supported for streaming. |
|
| C++ |
|
.NET is not supported for streaming. |
All operations and actions listed in this guide function equally across all supported SDKs.
Amazon Transcribe Service
Amazon Transcribe offers three main types of batch transcription: Standard, Medical, and Call Analytics.
-
Standard transcriptions are the most common option. Refer to StartTranscriptionJob for details.
-
Medical transcriptions are tailored to medical professionals and incorporate medical terms. A common use case for this service is transcribing doctor-patient dialogue into after-visit notes. Refer to StartMedicalTranscriptionJob for details.
-
Call Analytics transcriptions are designed for use with call center audio on two different channels; if you're looking for insight into customer service calls, use this option. Refer to StartCallAnalyticsJob for details.
Amazon Transcribe Streaming Service
Amazon Transcribe streaming offers four main types of real-time transcription: Standard, Medical, Call Analytics, and Health Scribe.
-
Standard transcriptions are the most common option. Refer to StartStreamTranscription for details.
-
Medical transcriptions are tailored to medical professionals and incorporate medical terms. A common use case for this service is transcribing doctor-patient dialogue in real time, so doctors can focus on their patient instead of taking notes. Refer to StartMedicalStreamTranscription for details.
-
Call Analytics transcriptions are designed for use with call center audio on two different channels; if you're looking for insight into customer service calls, use this option. Refer to StartCallAnalyticsStreamTranscription for details.
-
HealthScribe transcriptions are designed to automatically create clinical notes from patient-clinician conversations using generative AI. Refer to [here] for details.
