AgentPantheon
All Hands AI logo

All Hands AI开源的AI软件工程代理,自动化开发者工作流程。

4.5 (6)
Daniel Nikulshyn审阅者 Daniel Nikulshyn·更新 2026年7月

概览

All Hands AI 构建了开源自治代理,旨在承担真正的软件工程任务。代理可以读取代码库、编写和修改文件、在沙盒环境中执行命令、浏览网页,并与 API 交互,以与人类工程师共同完成开发工作。 最初名为 OpenDevin,该项目已发展为一个社区驱动的平台,既提供自托管的开源版本,也提供托管的云服务。它支持多种 LLM 后端,集成了 GitHub 等工具,面向想要透明、可定制的 AI 编码代理而非封闭专有助手的开发者。

主要功能

  • 具有沙盒执行的自主编码代理
  • 文件编辑、终端和浏览器工具使用
  • GitHub 集成和拉取请求工作流程
  • 可插拔的 LLM 后端(Anthropic、OpenAI、本地模型)
  • 自托管或托管云部署
  • 具有可扩展架构的开源代码库

价格

模型
Freemium
评分
4.5 / 5 (6)

使用场景

自动化拉取请求工作流程

将代理连接到 GitHub,以处理问题、实施修复和打开拉取请求,让工程师审查 AI 生成的更改,而不是从头开始编写样板代码。

自托管编码助手

在私有环境中部署开源代理,并选择您自己的 LLM 后端,为团队提供透明、可自定义的替代封闭专有编码助手。

端到端任务执行

交由多步骤开发任务,这些任务需要在沙盒环境中读取代码、编辑文件、运行终端命令和浏览文档。

实验 LLM 后端

在 Anthropic、OpenAI 或本地模型之间切换,以基准测试自主编码工作负载在真实存储库上的性能、成本和可靠性。

优点 & 缺点

优点

  • 完全开源且可自我托管
  • 支持多个 LLM 提供商
  • 积极的社区和快速发展
  • 处理端到端的编码任务,而不仅仅是建议

缺点

  • 自我托管需要技术设置
  • 代理可靠性因任务复杂度而异
  • 运行能力模型可能很昂贵
  • 仍在发展中,可能存在破坏性更改

评测

4.5

6 个评分的平均值。

5
3
4
3
3
0
2
0
1
0

登录以留下评测。

D

Devin Walker

Mar 15, 2026

Years in this space

I've evaluated a lot of these over the years. What stands out here is pluggable LLM backends (Anthropic, OpenAI, local models) — handled better than most — and supports multiple LLM providers. Still evolving with breaking changes is my one real gripe. Worth the time if this is your use case.

A

Aisha Khan

Dec 2, 2025

Does the job

Pretty happy overall. File editing, terminal, and browser tool use just works and handles end-to-end coding tasks, not just suggestions. Still evolving with breaking changes can be annoying, but no dealbreakers — I'd recommend it to a friend without hesitating.

S

Sanjay Gupta

Sep 22, 2025

Years in this space

I've evaluated a lot of these over the years. What stands out here is pluggable LLM backends (Anthropic, OpenAI, local models) — handled better than most — and handles end-to-end coding tasks, not just suggestions. Worth the time if this is your use case.

J

Joanna Kowalski

Sep 15, 2025

Skeptical, then convinced

I went in skeptical — most tools in this space overpromise. It actually delivers on autonomous coding agents with sandboxed execution, and fully open-source and self-hostable caught me off guard. still, I'd recommend giving it a real trial.

G

Gunnar Eriksson

Jul 1, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: autonomous coding agents with sandboxed execution and supports multiple LLM providers. Where it lags: agent reliability varies by task complexity. On balance the feature set — especially file editing, terminal, and browser tool use — justifies the 4 stars for our use case.

F

Fatima Zahra

May 30, 2025

Use it every day

Honestly didn't expect to like it this much. GitHub integration and pull request workflows is exactly what I needed, and supports multiple LLM providers. but I reach for it almost every day now and it just clicks.

问答

暂无问题 — 来当第一个提问的人吧。

提问

Software Development 的替代品