AWS Launches ‘Strands’: A Model Driven, Open Source SDK for Building AI Agents
With Strands, developers can simply define a prompt and a list of tools in code to build an agent, then test it locally and deploy it to the cloud.
AWS announced the release of Strands Agents, an open-source SDK designed to simplify the creation and deployment of AI agents.
The SDK adopts a model-driven approach, enabling developers to build agents with minimal coding, scaling from simple to complex use cases and from local development to production environments.
Strands Agents supports a variety of large language models (LLMs) through Amazon Bedrock, Anthropic’s Claude, Llama, Ollama, and OpenAI via LiteLLM, with the flexibility to integrate custom model providers.
It offers over 20 pre-built tools, including file manipulation, API requests, and AWS API interactions, and supports thousands of Model Context Protocol (MCP) servers for tool integration. Developers can also use Python functions as tools with a simple decorator.
The SDK has been utilized internally by AWS teams for services like Amazon Q Developer, AWS Glue, and VPC Reachability Analyzer, reducing development time from months to days or weeks. Strands Agents supports various architectures, including monolithic or microservices setups, and can be deployed for conversational agents, event-triggered tasks, or continuous operations.
Agents can be triggered by events, such as AWS Lambda functions responding to specific triggers (e.g., file uploads or API calls).
The project’s code is available on GitHub, with contributions from organizations like Accenture, Anthropic, Meta, and PwC. AWS aims to enhance developer productivity by replacing complex workflows with LLM-driven orchestration and plans to improve telemetry tooling in the future.
