AutoGen

situational

Microsoft framework for building multi-agent LLM applications.

Visit github.com →Add to kit Open in Noemium

ledger entry · observed by @whysanesanders

Multi-agent orchestration with human-in-the-loop, code execution, and group chat patterns. Still useful for research prototypes and complex reasoning chains, but now in maintenance mode and the debugging experience for agent conversations is rough.

Strengths

Open-source Microsoft framework with human-in-the-loop and code execution.
Supports group chat patterns for complex reasoning chains.
Free under the MIT license.

Known limitations

Now in maintenance mode; new users are directed to Microsoft Agent Framework.
Steep learning curve compared to CrewAI.
Documentation can lag the fast-moving codebase.
Production observability requires extra tooling.

Use it when

Research prototypes that need multi-agent conversation patterns.
Complex reasoning chains before committing to a managed platform.

Skip it when

You want active development rather than a maintenance-mode project.
Production observability and debugging are required out of the box.
The learning curve of a fast-moving codebase is a blocker.

Buy-check

·Price — free, open source (MIT) as of 2026-08-24. Open the pricing page, not a screenshot.
·Free tier — Free tier exists — burn it first.
·Cancel fallout — What happens to your data/projects on cancel — check the ToS, not the FAQ.
·Seat math — Per-seat pricing multiplies quietly. Count seats before the annual plan.
·Who skips it — Also skipped when: You want active development rather than a maintenance-mode project.; Production observability and debugging are required out of the box..