Why use this?

AutoGen is designed around agent conversations: multiple agents can exchange messages, call tools, execute code, and iterate toward a goal. This makes it useful for complex workflows that benefit from role separation, such as research, coding, planning, and operational automation.

Engineers choose AutoGen when they need fine-grained control over agent interactions and want a research-friendly framework with strong support for code execution and tool use.

Key Features

• Multi-agent conversations and role-based collaboration
• Tool calling and sandboxed code execution
• Human-in-the-loop workflows
• Support for multiple LLM providers
• Python-first APIs and extensible agent patterns