EXPLAINER

AutoGen (Microsoft)

frameworkfreshness: volatilekey claims checked 2026-10-09

AutoGen is an open-source Microsoft Research framework for multi-agent LLM applications where agents converse and collaborate. Its repository now states it is in maintenance mode, with Microsoft Agent Framework as the recommended successor for new projects.

Short answer: AutoGen popularised conversation-based multi-agent patterns (assistant, user-proxy, group chat). Its repository states it is in maintenance mode (no new features, community managed) and recommends Microsoft Agent Framework for new projects, with a migration guide for existing users. Status as of 2026-10-09; re-check before adopting.

What is it?

AutoGen lets several AI agents "talk" to each other and to humans to solve a task — one writes code, another reviews, a proxy runs it. It influenced many later frameworks. Because it has been redesigned and related Microsoft projects have changed, check what is current before starting.

Also known as: AutoGen, autogen studio, microsoft autogen, autogen agents, AG2, conversable agent, group chat agents, Microsoft Agent Framework.

Technical definition

The AutoGen paper (Wu et al., 2023) describes conversable agents and conversation programming. Subsequent major versions introduced a layered, event-driven redesign and tooling such as a visual studio; Microsoft also maintains related agent SDKs (see Semantic Kernel). As of 2026-10-09 the AutoGen repository states that AutoGen is in maintenance mode and that Microsoft Agent Framework (Python and .NET; repository github.com/microsoft/agent-framework) is its successor; the Semantic Kernel repository names the same successor. Existing AutoGen code keeps working but will not receive new features. Community forks with related names (for example AG2) are separate projects.

Entity type: framework · Freshness class: volatile.

When should I use it?

For new projects, evaluate Microsoft Agent Framework (the stated successor) rather than starting on AutoGen; keep AutoGen for maintaining existing systems or reading earlier research code.

Implementation patterns

  • Assistant + code-executor agents with sandboxed execution.
  • Group chat with a manager selecting speakers.
  • Human proxy for approvals.

Technologies and frameworks

Commonly associated: AutoGen (Python / .NET), AutoGen Studio, Docker for code execution. Names are examples of what exists, not recommendations, and projects change quickly.

Practical guide

  1. Check current repo status and docs.
  2. Decide whether conversational orchestration suits you.
  3. Sandbox code execution.
  4. Cap turns and tokens.
  5. Evaluate on tasks (How to Evaluate AI Systems and Outputs).

Common failure modes

  • Free-form agent chatter consuming tokens.
  • Executing generated code without a sandbox.
  • Following outdated tutorials for a different major version.
  • Unclear project lineage causing dependency confusion (forks/renames).

Troubleshooting

  • Examples do not run — Match the library major version to the tutorial; check package names.
  • Agents loop — Set max turns and termination conditions.

Contrasted with

  • LangGraph — LangGraph is graph-state-machine oriented; AutoGen centres on agent conversations/events.
  • CrewAI — CrewAI uses role/task crews; AutoGen uses conversational agent teams.

Questions this page answers

  • What is AutoGen?
  • AutoGen vs CrewAI vs LangGraph?
  • Is AutoGen still maintained?
  • What is the Microsoft Agent Framework?

Typical goals: evaluate AutoGen for multi-agent systems; work out which Microsoft agent framework is current.

Sources and verification

Status: key claims checked on 2026-10-09 against the primary sources listed below; everything else is not independently verified. The claims under "Checked against primary sources" were compared with the linked pages on that date. Pages change, so re-check anything time-sensitive before relying on it.

Checked against primary sources (2026-10-09):

  • The AutoGen README states AutoGen is in maintenance mode, will not receive new features or enhancements and is community managed.
  • It recommends Microsoft Agent Framework for new users and links the AutoGen migration guide; contributions are limited to bug fixes, security patches and documentation.
  • The AutoGen paper is arXiv:2308.08155 (resolved in the earlier citation audit).

Primary sources consulted:

Not independently verified:

  • Whether the maintenance-mode status changes later.
  • The relationship of community forks such as AG2 to Microsoft's project beyond the fact they are separate repositories.

Freshness class: volatile — Products, versions, protocols, benchmarks, leaderboards, regulations and dates. Verify against the primary source before relying on specifics; re-review every 3 months.

Time-sensitive claims to verify before relying on them:

  • Project status, forks (e.g. AG2), Microsoft agent-framework lineage and versions.
  • Package names and APIs.

Further reading and canonical sources (not all checked):

  • Wu et al., AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation (2023) — arXiv:2308.08155
  • AutoGen repository README (maintenance-mode notice) — github.com/microsoft/autogen
  • Microsoft Agent Framework repository — github.com/microsoft/agent-framework