@Jolyne_AI: 开源 AI 网页自动化工具:Nanobrowser。 OpenAI Operator 的开源替代方案,本地在浏览器里运行,支持多智能体协作。 免费、重视隐私、LLM 选择灵活、代码完全开源,让网页操作更智能、更高效。 GitHub:htt…

X AI KOLs Timeline 工具

摘要

Nanobrowser 是一个开源 AI 网页自动化工具,作为 OpenAI Operator 的免费替代方案,在本地浏览器中运行,支持多智能体协作,注重隐私且 LLM 选择灵活。

开源 AI 网页自动化工具:Nanobrowser。 OpenAI Operator 的开源替代方案,本地在浏览器里运行,支持多智能体协作。 免费、重视隐私、LLM 选择灵活、代码完全开源,让网页操作更智能、更高效。 GitHub:https://github.com/nanobrowser/nanobrowser… 从 GitHub 下载 zip,导入 Chrome 扩展即可开始使用。
查看原文
查看缓存全文

缓存时间: 2026/07/04 14:48

开源 AI 网页自动化工具:Nanobrowser。 OpenAI Operator 的开源替代方案,本地在浏览器里运行,支持多智能体协作。 免费、重视隐私、LLM 选择灵活、代码完全开源,让网页操作更智能、更高效。 GitHub:https://github.com/nanobrowser/nanobrowser… 从 GitHub 下载 zip,导入 Chrome 扩展即可开始使用。


nanobrowser/nanobrowser

Source: https://github.com/nanobrowser/nanobrowser

banner

GitHub Twitter Discord Ask DeepWiki Sponsor

🌐 Nanobrowser

Nanobrowser is an open-source AI web automation tool that runs in your browser. A free alternative to OpenAI Operator with flexible LLM options and multi-agent system.

⬇️ Get Nanobrowser from Chrome Web Store for free

👏 Join the community in Discord | X

🌟 Loving Nanobrowser? Give us a star and help spread the word!

❤️ Support the project by sponsoring us - every contribution helps keep Nanobrowser free and open source!

Nanobrowser Demo GIF

Nanobrowser's multi-agent system analyzing HuggingFace in real-time, with the Planner intelligently self-correcting when encountering obstacles and dynamically instructing the Navigator to adjust its approach—all running locally in your browser.

🔥Why Nanobrowser?

Looking for a powerful AI browser agent without the $200/month price tag of OpenAI Operator? Nanobrowser , as a chrome extension, delivers premium web automation capabilities while keeping you in complete control:

  • 100% Free - No subscription fees or hidden costs. Just install and use your own API keys, and you only pay what you use with your own API keys.
  • Privacy-Focused - Everything runs in your local browser. Your credentials stay with you, never shared with any cloud service.
  • Flexible LLM Options - Connect to your preferred LLM providers with the freedom to choose different models for different agents.
  • Fully Open Source - Complete transparency in how your browser is automated. No black boxes or hidden processes.

Note: We currently support OpenAI, Anthropic, Gemini, Ollama, Groq, Cerebras, Llama and custom OpenAI-Compatible providers, more providers will be supported.

📊 Key Features

  • Multi-agent System: Specialized AI agents collaborate to accomplish complex web workflows
  • Interactive Side Panel: Intuitive chat interface with real-time status updates
  • Task Automation: Seamlessly automate repetitive web automation tasks across websites
  • Follow-up Questions: Ask contextual follow-up questions about completed tasks
  • Conversation History: Easily access and manage your AI agent interaction history
  • Multiple LLM Support: Connect your preferred LLM providers and assign different models to different agents

🌐 Browser Support

Officially Supported:

  • Chrome - Full support with all features
  • Edge - Full support with all features

Not Supported:

  • Firefox, Safari, and other Chromium variants (Opera, Arc, etc.)

Note: While Nanobrowser may function on other Chromium-based browsers, we recommend using Chrome or Edge for the best experience and guaranteed compatibility.

🚀 Quick Start

  1. Install from Chrome Web Store (Stable Version):

Important Note: For latest features, install from “Manually Install Latest Version” below, as Chrome Web Store version may be delayed due to review process.

  1. Configure Agent Models:
    • Click the Nanobrowser icon in your toolbar to open the sidebar
    • Click the Settings icon (top right)
    • Add your LLM API keys
    • Choose which model to use for different agents (Navigator, Planner)

🔧 Manually Install Latest Version

To get the most recent version with all the latest features:

  1. Download

    • Download the latest nanobrowser.zip file from the official Github release page.
  2. Install:

    • Unzip nanobrowser.zip.
    • Open chrome://extensions/ in Chrome
    • Enable Developer mode (top right)
    • Click Load unpacked (top left)
    • Select the unzipped nanobrowser folder.
  3. Configure Agent Models

    • Click the Nanobrowser icon in your toolbar to open the sidebar
    • Click the Settings icon (top right).
    • Add your LLM API keys.
    • Choose which model to use for different agents (Navigator, Planner)
  4. Upgrading:

    • Download the latest nanobrowser.zip file from the release page.
    • Unzip and replace your existing Nanobrowser files with the new ones.
    • Go to chrome://extensions/ in Chrome and click the refresh icon on the Nanobrowser card.

🛠️ Build from Source

If you prefer to build Nanobrowser yourself, follow these steps:

  1. Prerequisites:

    • Node.js (v22.12.0 or higher)
    • pnpm (v9.15.1 or higher)
  2. Clone the Repository:

    git clone https://github.com/nanobrowser/nanobrowser.git
    cd nanobrowser
    
  3. Install Dependencies:

    pnpm install
    
  4. Build the Extension:

    pnpm build
    
  5. Load the Extension:

    • The built extension will be in the dist directory
    • Follow the installation steps from the Manually Install section to load the extension into your browser
  6. Development Mode (optional):

    pnpm dev
    

🤖 Choosing Your Models

Nanobrowser allows you to configure different LLM models for each agent to balance performance and cost. Here are recommended configurations:

Better Performance

  • Planner: Claude Sonnet 4
    • Better reasoning and planning capabilities
  • Navigator: Claude Haiku 3.5
    • Efficient for web navigation tasks
    • Good balance of performance and cost

Cost-Effective Configuration

  • Planner: Claude Haiku or GPT-4o
    • Reasonable performance at lower cost
    • May require more iterations for complex tasks
  • Navigator: Gemini 2.5 Flash or GPT-4o-mini
    • Lightweight and cost-efficient
    • Suitable for basic navigation tasks

Local Models

  • Setup Options:

    • Use Ollama or other custom OpenAI-compatible providers to run models locally
    • Zero API costs and complete privacy with no data leaving your machine
  • Recommended Models:

  • Prompt Engineering:

    • Local models require more specific and cleaner prompts
    • Avoid high-level, ambiguous commands
    • Break complex tasks into clear, detailed steps
    • Provide explicit context and constraints

Note: The cost-effective configuration may produce less stable outputs and require more iterations for complex tasks.

Tip: Feel free to experiment with your own model configurations! Found a great combination? Share it with the community in our Discord to help others optimize their setup.

💡 See It In Action

Here are some powerful tasks you can accomplish with just a sentence:

  1. News Summary:

    “Go to TechCrunch and extract top 10 headlines from the last 24 hours”

  2. GitHub Research:

    “Look for the trending Python repositories on GitHub with most stars”

  3. Shopping Research:

    “Find a portable Bluetooth speaker on Amazon with a water-resistant design, under $50. It should have a minimum battery life of 10 hours”

🛠️ Roadmap

We’re actively developing Nanobrowser with exciting features on the horizon, welcome to join us!

Check out our detailed roadmap and upcoming features in our GitHub Discussions.

🤝 Contributing

We need your help to make Nanobrowser even better! Contributions of all kinds are welcome:

  • Share Prompts & Use Cases
    • Join our Discord server.
    • share how you’re using Nanobrowser. Help us build a library of useful prompts and real-world use cases.
  • Provide Feedback
    • Try Nanobrowser and give us feedback on its performance or suggest improvements in our Discord server.
  • Contribute Code
    • Check out our CONTRIBUTING.md for guidelines on how to contribute code to the project.
    • Submit pull requests for bug fixes, features, or documentation improvements.

We believe in the power of open source and community collaboration. Join us in building the future of web automation!

🔒 Security

If you discover a security vulnerability, please DO NOT disclose it publicly through issues, pull requests, or discussions.

Instead, please create a GitHub Security Advisory to report the vulnerability responsibly. This allows us to address the issue before it’s publicly disclosed.

We appreciate your help in keeping Nanobrowser and its users safe!

💬 Community

Join our growing community of developers and users:

👏 Acknowledgments

Nanobrowser builds on top of other awesome open-source projects:

Huge thanks to their creators and contributors!

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Made with ❤️ by the Nanobrowser Team.

Like Nanobrowser? Give us a star 🌟 and join us in Discord | X

⚠️ DISCLAIMER ON DERIVATIVE PROJECTS

We explicitly DO NOT endorse, support, or participate in any projects involving cryptocurrencies, tokens, NFTs, or other blockchain-related applications based on this codebase.

Any such derivative projects are NOT Affiliated with, or maintained by, or in any way connected to the official Nanobrowser project or its core team.

We assume NO LIABILITY for any losses, damages, or issues arising from the use of third-party derivative projects. Users interact with these projects at their own risk.

We reserve the right to publicly distance ourselves from any misuse or misleading use of our name, codebase, or brand.

We encourage open-source innovation but urge our community to be discerning and cautious. Please ensure you understand the risks before using any software or service built upon our codebase by independent developers.

相似文章

@GitHub_Daily: 让 AI Agent 自动化操作浏览器或抓数据,经常被各种反爬机制拦截,遇到验证码、人机验证直接卡死。 最近 BrowserAct 团队开源了一个 Skill,专为 AI Agent 设计的浏览器自动化命令行工具。 提供三层反封锁机制,从…

X AI KOLs Timeline

BrowserAct 团队开源了一个专为 AI Agent 设计的浏览器自动化命令行工具,提供三层反封锁机制(指纹伪装、验证码破解、人类接管),支持多浏览器并行、账户隔离,并优化了输出格式以节省Token。

Open Browser Use

Product Hunt

Open Browser Use 是一个面向本地AI代理的开源浏览器自动化工具。

@wsl8297: 在 GitHub 挖到一个「100% 本地」的 Manus AI 开源替代方案:AgenticSeek。 它把本地推理模型和智能代理系统打包在一起:能自己逛网页、搜资料、填表单,能写代码、做规划,还支持语音对话。整体体验很接近电影里 Ja…

X AI KOLs Timeline

AgenticSeek 是一个 100% 本地的 Manus AI 开源替代方案,集成了本地推理模型和智能代理系统,支持网页浏览、编程、语音交互等,数据完全存储在本地。

介绍 Operator

OpenAI Blog

OpenAI 发布了 Operator,一款 AI 代理,可以自主使用网络浏览器执行表单填写、在线购物和网络交互等任务。目前作为研究预览版向美国 Pro 用户开放,计划进一步扩展并集成到 ChatGPT 中,由新推出的计算机使用代理(CUA)模型驱动。