---
title: "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation"
license: CC-BY-4.0
license_url: https://creativecommons.org/licenses/by/4.0/
source: https://arxiv.org/abs/2308.08155
source_license: "CC BY 4.0"
attribution: "Summary by iinventedlight (AI-assisted). Source: arXiv, \"AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation\" (CC BY 4.0)."
condensed_by: "iinventedlight (Grok Bot, AI-assisted)"
fact_checked: 2026-11-02
date: 2026-11-02
status: reviewed
reviewed_on: 2026-11-02
category: arxiv
arxiv_id: 2308-08155-autogen
alias: 2308-08155-autogen
checked_on: 2026-11-02
license_url: https://creativecommons.org/licenses/by/4.0/
source: https://arxiv.org/abs/2308.08155
source_license: "CC BY 4.0 (arXiv paper)"
arxiv_license_url: https://creativecommons.org/licenses/by/4.0/
arxiv_license_checked: 2026-11-02
arxiv_license_check_note: "License link read from the paper\'s own arXiv abstract page on 2026-11-02."
summary_policy: "Short original summary with attribution; no passages copied, at most brief phrases quoted."
attribution: "Summary by iinventedlight (AI-assisted), condensed and paraphrased; sources: Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, et al., \"AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation\", arXiv:2308.08155 (CC BY 4.0)"
condensed_by: "iinventedlight (Grok Bot, AI-assisted)"
fact_checked: 2026-11-02
date: 2026-11-02
category: arxiv
tags: arxiv, llm, paper-summary
checked_on: 2026-11-02
sources:
  - id: S1
    title: "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation"
    authors: "Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, et al."
    publisher: "arXiv"
    url: https://arxiv.org/abs/2308.08155
    accessed: 2026-11-02
    license: "CC BY 4.0"
    attribution: "Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, et al., \"AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation\", arXiv:2308.08155 (CC BY 4.0)"
---
# AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation

# AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation

## Summary
AutoGen is an open-source framework for building LLM applications from several agents that talk to each other. Agents are customizable and conversable, and they can combine LLMs, human input and tools. Developers program the conversation patterns in natural language or code. The paper shows example applications in math, coding, question answering, operations research and more.

*Draft. License: CC BY 4.0 (this summary). Sources: [S1] arXiv, "AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation" (CC BY 4.0).*

## When to apply
- Splitting a task into roles (e.g. writer + code executor + reviewer), with a human able to step in.

## Paper
- Title: AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
- Authors: Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, et al.
- arXiv: https://arxiv.org/abs/2308.08155 (first posted 2023-08-16)
- License: CC BY 4.0 (checked on the arXiv abstract page 2026-11-02)
- Read for this note: arXiv abstract page; figures restate the abstract.

## Facts
- id: F1
  claim: AutoGen is an open-source framework for building LLM applications from multiple agents that converse to accomplish tasks.
  confidence: medium
  date: 2026-11-02
  sources: [S1]
  license: CC BY 4.0
  checked_on: 2026-11-02
- id: F2
  claim: Agents can operate in modes combining LLMs, human inputs and tools.
  confidence: medium
  date: 2026-11-02
  sources: [S1]
  license: CC BY 4.0
  checked_on: 2026-11-02
- id: F3
  claim: Interaction behavior and conversation patterns can be programmed in natural language and in code.
  confidence: medium
  date: 2026-11-02
  sources: [S1]
  license: CC BY 4.0
  checked_on: 2026-11-02
- id: F4
  claim: Example applications span mathematics, coding, question answering, operations research, online decision-making and entertainment (author-reported).
  confidence: medium
  date: 2026-11-02
  sources: [S1]
  license: CC BY 4.0
  checked_on: 2026-11-02

## Caveats
- The abstract reports effectiveness qualitatively; check the paper for specific numbers.
- The framework has changed a lot since 2023; check current docs before using APIs named in the paper.

## Pointers
- id: P1
  answers: Full method, experiments and appendices
  where: https://arxiv.org/abs/2308.08155
  access: free, public web page