Strands Agents

Strands Agents is an open-source Python SDK developed by Amazon Web Services that provides a model-driven approach to building AI agents. Unlike rigid framework-based approaches, Strands lets the underlying language model dynamically decide which tools to use and in what order, making agent behavior more natural and adaptive. The SDK supports multiple LLM providers including Amazon Bedrock, Anthropic, LiteLLM, Ollama, and any OpenAI-compatible endpoint, giving developers flexibility in their model choice.

Strands provides built-in tools for file operations, shell commands, HTTP requests, code execution, RAG retrieval, and more. Developers can also create custom tools with a simple decorator pattern. The framework includes native conversation memory management, allowing agents to maintain context across multi-turn interactions without complex state management code.

One of the key strengths of Strands is its simplicity — a basic agent can be created in just three lines of code, while still supporting complex multi-agent orchestration patterns for production systems. The SDK integrates seamlessly with AWS services like Bedrock for model hosting and Lambda for serverless deployment, but is not locked into the AWS ecosystem. It also supports autonomous agent loops where the agent can plan, execute, and iterate on complex tasks with minimal human intervention. Strands has quickly gained traction in the agent development community as a lightweight but powerful alternative to heavier frameworks like LangChain and CrewAI.