Glossary
AutoGen
A Microsoft Research framework for multi-agent systems, with a conversation-pattern API for orchestrating multiple specialized agents to solve tasks collaboratively.
Microsoft Research’s framework for building multi-agentagentsArchitectures where multiple LLM-driven agents collaborate or compete on a task, each with its own role, prompt, or specialization, coordinated by an orchestrator or message-passing protocol. Open full entry LLM systems. AutoGen abstracts agent interaction as a conversation: a UserProxy agent passes work to an Assistant agent or a GroupChat of multiple specialized agents; the framework manages turn-taking, tool useagentsThe general pattern of an LLM invoking external functions, APIs, or systems to fetch data or take action, distinct from generating an answer purely from its weights. Open full entry , and shared state.
The release came at a moment of high enthusiasm for multi-agentagentsArchitectures where multiple LLM-driven agents collaborate or compete on a task, each with its own role, prompt, or specialization, coordinated by an orchestrator or message-passing protocol. Open full entry patterns. By 2026 the empirical case for multi-agent is contested (some research argues for, the Cognition essay argues against); AutoGen remains a useful reference implementation of the patterns either way. MIT licensed.
Full coverage at /projects/autogen.