Home · Frameworks · AutoGen
FRAMEWORK

AutoGen

Multi-agent conversation framework enabling complex LLM workflows through agent-to-agent communication.

TARGET QUERY autogen microsoft · ~15K/mo
CREATOR
Microsoft
LANGUAGE
Python
CATEGORY
Orchestration
GITHUB STARS
~32K
LICENSE
MIT
OVERVIEW Updated 2026-05-17

What It Does

AutoGen is Microsoft’s framework for building multi-agent conversational AI systems. Agents communicate through natural language messages, can execute code, use tools, and collaborate to solve complex tasks through iterative conversation.

Key Features

AutoGen offers customizable conversational agents, code execution in sandboxed environments, group chat for multi-agent conversations, human proxy agents for human-in-the-loop, nested conversations, teachable agents that learn from interactions, and flexible conversation patterns.

When To Use It

Use AutoGen when you need agents that can write and execute code iteratively, want flexible multi-agent conversations rather than fixed workflows, or need human oversight integrated into agent communication. It excels at coding tasks and iterative problem solving.

Alternatives

CrewAI provides more structured role-based workflows. LangGraph offers finer control over agent state. Semantic Kernel targets enterprise C# and Python integration. For simple agent tasks, LangChain agents are lighter weight.