Model Context Protocol (MCP) Deployment Patterns on AWS
Deepika Kumar and Lingling Cui, Amazon Web Services
Executive Summary
The Model Context Protocol (MCP) is an open protocol that enables seamless integration between AI applications and external data sources and tools. As organizations increasingly adopt AI-powered development tools like Kiro, Amazon Q Developer, Claude Code, and custom AI assistants, deploying MCP servers on AWS provides scalable, secure, and cost-effective infrastructure for contextual AI interactions.
This prescriptive guidance provides architecture patterns and implementation approaches for deploying MCP servers on AWS using AWS Lambda with Amazon API Gateway, Amazon Elastic Container Service (Amazon ECS), and Amazon Elastic Kubernetes Service (Amazon EKS) and others. Each pattern addresses different operational requirements, scaling needs, and management preferences.
The implementation examples and deployment scripts referenced in this guidance are available in the open-source repository at: https://github.com/aws-samples/sample-mcp-deployment-patterns
Intended audience
This guide is intended for cloud architects, DevOps engineers, platform engineers, and developers who want to deploy MCP servers on AWS infrastructure. It assumes L200-300 familiarity with core AWS services, networking concepts (Amazon VPC, subnets, security groups), and container fundamentals.
Objectives
After reading this guide, you will be able to:
Understand the differences between local and remote MCP server deployment models
Evaluate five AWS deployment patterns (Amazon Bedrock AgentCore, Lambda + API Gateway, Amazon ECS, Amazon EKS, Amazon EC2) for hosting remote MCP servers
Select the appropriate deployment pattern based on your scaling, security, cost, and operational requirements
Apply best practices for production MCP deployments aligned with the AWS Well-Architected Framework
Attachments
To access additional content that is associated with this document, download and unzip the following file:
