Implementing programmatic tool calling on Amazon Bedrock
Three ways to run tool calling on Bedrock—from Docker sandbox to managed agents. Pick your control vs. convenience tradeoff.
Why it matters
AWS is making it easier for enterprises to deploy agentic AI workflows on Bedrock by offering multiple implementation paths (self-hosted, managed, SDK-compatible), lowering the friction for teams building production AI applications.
The key facts
9 to knowThree implementation paths for programmatic tool calling on Amazon Bedrock
Self-hosted Docker sandbox option on ECS for maximum control
Managed solution via Amazon Bedrock AgentCore Code Interpreter
Anthropic SDK-compatible proxy path for developer experience parity
Published May 19, 2026 on AWS ML blog
Three implementation paths: self-hosted Docker/ECS sandbox, managed Bedrock AgentCore Code Interpreter, Anthropic SDK-compatible proxy
Targets developer experience parity with Anthropic SDK
Reduces migration friction for Claude users on AWS infrastructure
Enables programmatic function calling without vendor lock-in options
Go to the source
AWS Machine Learning Blogaws.amazon.com
Publisher excerpt: In this post, we show three ways to implement Programmatic tool calling (PTC) on Amazon Bedrock: a self-hosted Docker sandbox on ECS for maximum control, a managed solution using Amazon Bedrock AgentCore Code Interpreter, and an Anthropic SDK-compatible path through a proxy for teams that prefer…
