项目介绍
Agent 智能体 · 2026-11-09 更新 · 本站尚未实测
Open Interpreter 目前是一个面向低成本模型优化的编码代理(coding agent),并提供终端交互界面。按 README 说明,它是 OpenAI Codex 的一个分支(fork),核心目标是“模拟最能发挥低成本模型性能的 agent harness”,并以 Rust 重写。
需要特别区分版本:仓库 README 明确提示,这是“新的 Rust 版本”,早期 Python 版项目作为社区维护分支存在于 endolith/open-interpreter。查找旧版 Python 用法时不应套用到本仓库。
核心能力
点击能力名称查看输入、输出和原文依据。
- 终端代理会话:安装后在终端输入
i或interpreter启动会话。 - Harness 切换:使用
/harness在 native、claude-code、zcode、kimi-code、kimi-cli、qwen-code、deepseek-tui、swe-agent、minimal 等 harness 之间切换。 - 模型与供应商切换:通过
/model在 TUI 中切换;可通过interpreter --chat-completions或interpreter exec --chat-completions走任意 OpenAI 兼容供应商的 Chat Completions。 - ACP 兼容:可被 ACP 兼容编辑器与客户端调用,命令为
interpreter acp。 - Codex 兼容:README 称其遵循相同的 Codex exec 协议,并提供
codexPathOverride: "interpreter"的一行替换方式接入 Codex SDK。 - 共享格式:支持仓库内
AGENTS.md、.agents/skills目录、MCP、ACP、hooks、权限等。 - 计算机操作与测试:内置 QA 技能,可用 agent-browser 操作真实浏览器中的 Web 应用,或用 trycua 操作与测试原生应用。
- 沙箱:README 称可在 macOS、Linux、Windows 的原生沙箱中运行命令。
实际应用场景
适用场景
- 希望在终端中以对话方式完成编码与命令执行任务的开发者。
- 使用低成本或开源权重模型(README 提到 Kimi K3、GLM、DeepSeek 等供应商文档)并希望借助 harness 调优效果的团队。
- 已在 Codex SDK 或 ACP 客户端上构建工具链,想替换底层代理可执行文件的场景。
- 需要让模型驱动浏览器或原生应用做界面测试的质量保障流程。
安装与接入
README 给出的安装方式为脚本安装。macOS 与 Linux 使用 curl -fsSL https://www.openinterpreter.com/install | sh;Windows 使用 PowerShell 的 irm https://www.openinterpreter.com/install.ps1 | iex。安装后在终端输入 i 或 interpreter 启动会话。
接入 ACP 客户端时,将客户端配置为启动 interpreter acp,具体示例见官方 ACP 文档。使用 Codex SDK 时按 README 方式指定可执行文件路径为 interpreter。
模型供应商的配置步骤、可用模型清单与凭证要求,README 指向官方 provider 文档,本页未逐一列出,具体配置请以官方文档为准。
使用条件与限制
- 是否需要部署
- 需自行准备运行环境,具体要求见正文
- 是否需要注册
- 账号和认证要求见正文;未明确说明的项目仍待核验
- 是否需要 API Key
- 待核验
- 费用与额度
- 费用边界见正文,项目许可不代表模型或托管服务免费
- 所需权限
- 待核验
- 许可证与服务条款
- 待核验
权限费用与限制
- 权限:代理会在本机运行命令,README 提到原生沙箱与审批机制(sandbox & approvals 文档),但具体默认策略与授权范围未在 README 中展开,需查阅官方文档并自行评估。
- 费用:软件以 Apache-2.0 发布,但模型调用通常由第三方供应商计费;是否免费、是否有商业模块或托管服务价格,资料未说明,待向官方核验。
- 限制:不同 harness 的兼容性与效果差异、各供应商模型的支持范围、Windows 具体功能完整度,README 未给出完整对照,需实测或查官方文档确认。
- 旧版差异:Python 版与当前 Rust 版命令、配置可能不同,不要混用。
国内使用情况
网友核验 用户经验分享
与站长实测独立,不代表官方认证或本站已实测。
最近核验:暂无有效核验。参与人数按当前项目有效身份去重;不同核验项分别统计,不合并成用户总结果。仅统计审核通过且未过期的记录。
分项统计、有效期与样本量
| 核验项目 | 有效人数 | 成功 / 部分 / 失败 / 无法判断 | 成功率 | 最近核验 / 有效期 |
|---|---|---|---|---|
| 官方网站访问 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 7 天 |
| GitHub 仓库访问 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 7 天 |
| 安装与运行 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 官方服务国内连接 / MCP 连接 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 国内网络情况 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 7 天 |
| 国内模型兼容性 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 自定义 API | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 海外服务依赖 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 海外账号要求 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 中文界面 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 客户端兼容性 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
| 账号及费用信息 | 0 | 0 / 0 / 0 / 0 | 样本不足 | 暂无 / 30 天 |
成功率 = 成功 ÷(成功 + 部分成功 + 失败);无法判断不计入分母。至少 5 份可判断结果才显示成功率。账号、费用、海外依赖与语言等事实项,成功表示核查结论与资料吻合,具体观察另行统计;不表示免费或没有海外依赖。不同网络、客户端和版本的结果不能直接等同。
最近有效核验
暂无审核通过且未过期的网友核验。欢迎提交真实使用记录。
查看全部有效核验 →国内实测状态:国内可用性待核验。协议兼容和官方资料说明不等于国内网络可用。
- 官方网站访问
- 尚未实测 待核验
- GitHub 仓库访问
- 尚未实测 待核验
- 官方服务国内连接
- 尚未实测 待核验
- 国内网络情况
- 尚未实测 待核验
- 国内模型兼容性
- 尚未实测 待核验
- 自定义 API
- 尚未实测 待核验
- 海外服务依赖
- 尚未实测 待核验
- 海外账号要求
- 尚未实测 待核验
- 中文界面
- 尚未实测 待核验
- 国内最后核验时间
- 尚未实测
许可与国内使用
仓库代码采用 Apache-2.0 许可,LICENSE 文件版权标注为 2025 OpenAI。该许可仅覆盖仓库中的代码,不自动覆盖第三方模型、API 服务、托管服务或内置技能所依赖的外部项目(如 agent-browser、trycua)。
README 提供简体中文版说明页,并列出 Kimi、DeepSeek、Z.AI/GLM 等供应商文档,但未说明各服务在中国大陆的网络可达性、账号注册或合规要求;国内可用性需按所选供应商单独核验。
项目资料与来源
字段依据与待核验事项
access:来源确认 · Then type `i` or `interpreter` in your terminal to start a session.
project_type:来源确认 · A coding agent optimized for low-cost models.
原始 README(折叠技术参考)
<h1 align="center">Open Interpreter</h1>
<p align="center">A coding agent optimized for low-cost models. <a href="https://www.openinterpreter.com/blog/open-interpreter?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=hero_text"><strong>Blog post ↗</strong></a></p>
<p align="center">
<b>English</b> • <a href="README_ES.md">Español</a> • <a href="README_ZH.md">简体中文</a> • <a href="README_JA.md">日本語</a>
</p>
<p align="center">
<a href="https://discord.gg/Hvz9Axh84z"><img alt="Discord" src="https://img.shields.io/discord/1146610656779440188?style=flat-square&label=Discord" /></a>
<a href="https://www.openinterpreter.com/docs/terminal?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=docs_badge"><img alt="Documentation" src="https://img.shields.io/badge/Documentation-white?style=flat-square" /></a>
<a href="LICENSE"><img alt="License" src="https://img.shields.io/badge/License-Apache--2.0-white?style=flat-square" /></a>
</p>
> [!NOTE]
> **Today: Kimi K3 is here.** We have reimplemented the provider-recommended
> [Kimi Code](https://www.kimi.com/coding/en) harness in Rust, giving you
> maximum K3 performance with a Codex-like interface.
> [**Kimi Docs →**](https://www.openinterpreter.com/docs/terminal/kimi-k3?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=kimi_k3_note)
<br>
<p align="center">
<a href="https://www.openinterpreter.com/blog/open-interpreter?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=hero_image">
<img alt="Open Interpreter running in a terminal" src="docs-site/assets/open-interpreter-terminal-hero.png" width="100%" />
</a>
</p>
## Installation
macOS and Linux:
```bash
curl -fsSL https://www.openinterpreter.com/install | sh
```
Windows:
```powershell
irm https://www.openinterpreter.com/install.ps1 | iex
```
Then type `i` or `interpreter` in your terminal to start a session.
## Harness Emulation
Open Interpreter is a fork of OpenAI's Codex, with a focus on emulating the agent harness that gets the best performance out of low-cost models.
Use `/harness` to switch the active harness:
```text
> /harness
native
claude-code
claude-code-bare
zcode
kimi-code
kimi-cli
qwen-code
deepseek-tui
swe-agent
minimal
```
Read more in the [harness docs](https://www.openinterpreter.com/docs/terminal/harness?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=harness_docs) and [provider setup guides](https://www.openinterpreter.com/docs/terminal/providers?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=provider_guides).
## ACP compatible, Codex compatible
Open Interpreter works in [ACP-compatible editors and clients](https://agentclientprotocol.com/get-started/clients). Configure the client to launch `interpreter acp`; see the [ACP guide](https://www.openinterpreter.com/docs/terminal/acp) for examples.
Already building with OpenAI's Codex SDK? Keep the SDK and make a one-line
binary override:
```diff
-const codex = new Codex();
+const codex = new Codex({ codexPathOverride: "interpreter" });
```
Open Interpreter speaks the same Codex exec protocol. See the [SDK guide](https://www.openinterpreter.com/docs/terminal/sdk) and run `scripts/test-codex-sdk-compat.sh` for a local, provider-free compatibility check.
Want a small browser UI over raw Open Interpreter? Start with the
[local streaming web chat guide](docs/streaming-web-chat.md). For a complete
desktop, browser, or headless client, use
[Interpreter Workstation](https://github.com/openinterpreter/interpreter-workstation).
## Portable by default
Open Interpreter should fit into your existing agent setup instead of trapping
it in an Open Interpreter-only format. The product goal is to prefer shared,
tool-neutral standards and directories, keep user-authored data in readable
files, and make moving to or from another compatible agent straightforward.
Today that includes repository `AGENTS.md`, shared `.agents/skills` directories,
MCP, ACP, and the Codex exec protocol. Product-specific storage under
`~/.openinterpreter` is reserved for configuration and runtime state that does
not yet have a practical shared standard. Legacy product-specific skill
directories remain readable for compatibility, but new skills belong in
`.agents/skills` or `~/.agents/skills`.
See the [portability guide](docs/portability.md) for the current boundary and
the rules for evolving it.
## Computer Use
Open Interpreter ships with a QA skill that lets any model operate and test interfaces. It can drive web apps in a real browser with [agent-browser](https://github.com/vercel-labs/agent-browser), or operate and test native apps with [trycua](https://github.com/trycua/cua).
## Features
- Runs commands inside native sandboxing on macOS, Linux, and Windows.
- Switches providers and models from the TUI with `/model`.
- Runs any selected OpenAI-compatible provider through Chat Completions with
`interpreter --chat-completions` or `interpreter exec --chat-completions`.
- Inspects or switches Rust-native model harnesses with `/harness`.
- Tests web and native apps through the built-in QA skill.
- Runs as an [Agent Client Protocol](https://agentclientprotocol.com/) agent for editors with `interpreter acp`.
- Reuses shared `AGENTS.md` instructions and `.agents/skills` directories.
- Keeps product-only config and session state local under `~/.openinterpreter`.
- Supports `exec`, MCP, skills, hooks, permissions, and `AGENTS.md`.
## Documentation
- [Terminal docs](https://www.openinterpreter.com/docs/terminal?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=terminal_docs)
- [Quickstart](https://www.openinterpreter.com/docs/terminal/quickstart?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=quickstart)
- [Install guide](https://www.openinterpreter.com/docs/terminal/install?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=install_guide)
- [Configuration](https://www.openinterpreter.com/docs/terminal/config?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=configuration)
- [CLI reference](https://www.openinterpreter.com/docs/terminal/cli-reference?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=cli_reference)
- [Harnesses](https://www.openinterpreter.com/docs/terminal/harness?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=harnesses)
- [Model provider guides](https://www.openinterpreter.com/docs/terminal/providers?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=provider_guides)
- [Kimi K3](https://www.openinterpreter.com/docs/terminal/kimi-k3?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=kimi_k3_docs)
- [DeepSeek](https://www.openinterpreter.com/docs/terminal/deepseek?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=deepseek_docs)
- [Z.AI, GLM, and ZCode](https://www.openinterpreter.com/docs/terminal/zai-glm?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=zai_glm_docs)
- [Agent Client Protocol](https://www.openinterpreter.com/docs/terminal/acp)
- [Codex SDK](https://www.openinterpreter.com/docs/terminal/sdk)
- [Portability](https://github.com/openinterpreter/openinterpreter/blob/main/docs/portability.md)
- [Sandbox & approvals](https://www.openinterpreter.com/docs/terminal/sandbox?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=sandbox_approvals)
- [Branding a distribution fork](FORK_BRANDING.md)
Provider and model membership is generated, not maintained as Rust lists. From
`codex-rs`, refresh all hosted providers with
`python3 scripts/write_provider_catalog.py`, or repeat
`--provider <provider-id>` to update only selected provider entries. Live model
sources require the provider credentials documented in the
[provider docs](https://www.openinterpreter.com/docs/terminal/providers?utm_source=github&utm_medium=referral&utm_campaign=readme&utm_content=provider_catalog_generation).
> [!NOTE]
> This is the new Rust version of Open Interpreter, based on Codex. Looking for the original Python project? It lives on as a community-maintained fork at [endolith/open-interpreter](https://github.com/endolith/open-interpreter).
## License
Apache-2.0
历史前置条件说明(补充参考)
需 macOS、Linux 或 Windows(README 声明支持三者,具体功能完整度待核验);通过官方安装脚本获取,需可访问 openinterpreter.com;使用模型需配置相应供应商凭证或 API Key,具体变量与步骤待核验;模型调用可能产生第三方费用,价格与免费额度待核验;国内网络与账号可用性待核验。
项目来源与清单版权许可
https://github.com/OpenInterpreter/open-interpreter
整理来源:https://github.com/slavakurilyak/awesome-ai-agents
清单版权与许可:
MIT License
Copyright (c) 2024 Slava Kurilyak
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.