PluginBench
MCP Server
Active
Apache-2.0

io.github.bytedance/mcp-server-browser MCP Server

io.github.bytedance/mcp-server-browser

Browser automation for AI agents via MCP, powering ByteDance's Agent TARS hybrid GUI/DOM browser control.

RepositoryRegistry

What is the io.github.bytedance/mcp-server-browser MCP server?

The mcp-server-browser (@agent-infra/mcp-server-browser) is an MCP server from ByteDance's UI-TARS-desktop project that exposes browser automation capabilities to AI agents via the Model Context Protocol. It lets an MCP client (like Claude or Cursor) control a web browser—navigating pages, interacting with DOM elements, and performing GUI-based actions—as part of a broader multimodal agent stack (Agent TARS / UI-TARS Desktop). It is designed to be mounted as an MCP tool so agents can complete real-world browser tasks like booking flights or hotels.

mcp-server-browser is part of ByteDance's UI-TARS-desktop monorepo, which ships the Agent TARS multimodal AI agent stack and the UI-TARS Desktop GUI agent application. This particular package provides browser-use access as an MCP server, allowing AI agents to control browsers through a hybrid strategy combining GUI Agent (visual grounding/screenshots) and DOM-based interaction. It's used as one of the pluggable MCP tools that Agent TARS mounts to let LLM-driven agents perform real-world tasks such as booking flights/hotels or navigating websites, driven by an Event Stream protocol for context engineering.

How to install io.github.bytedance/mcp-server-browser

Copy-paste configuration for popular MCP clients.

transport: stdio
Config generated by PluginBench — verify against the source before use.
Environment / auth
  • DISPLAY

    DISPLAY environment variable for browser rendering

Claude Desktop
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "mcp-server-browser": {
      "command": "npx",
      "args": [
        "-y",
        "@agent-infra/mcp-server-browser",
        "browser",
        "cdp-endpoint",
        "ws-endpoint",
        "executable-path",
        "output-dir",
        "proxy-bypass",
        "proxy-server",
        "vision"
      ],
      "env": {
        "DISPLAY": "<YOUR_DISPLAY>"
      }
    }
  }
}
Cursor
~/.cursor/mcp.json
{
  "mcpServers": {
    "mcp-server-browser": {
      "command": "npx",
      "args": [
        "-y",
        "@agent-infra/mcp-server-browser",
        "browser",
        "cdp-endpoint",
        "ws-endpoint",
        "executable-path",
        "output-dir",
        "proxy-bypass",
        "proxy-server",
        "vision"
      ],
      "env": {
        "DISPLAY": "<YOUR_DISPLAY>"
      }
    }
  }
}
Windsurf
~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "mcp-server-browser": {
      "command": "npx",
      "args": [
        "-y",
        "@agent-infra/mcp-server-browser",
        "browser",
        "cdp-endpoint",
        "ws-endpoint",
        "executable-path",
        "output-dir",
        "proxy-bypass",
        "proxy-server",
        "vision"
      ],
      "env": {
        "DISPLAY": "<YOUR_DISPLAY>"
      }
    }
  }
}
VS Code
.vscode/mcp.json
{
  "servers": {
    "mcp-server-browser": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@agent-infra/mcp-server-browser",
        "browser",
        "cdp-endpoint",
        "ws-endpoint",
        "executable-path",
        "output-dir",
        "proxy-bypass",
        "proxy-server",
        "vision"
      ],
      "env": {
        "DISPLAY": "<YOUR_DISPLAY>"
      }
    }
  }
}
Claude Code
claude mcp add mcp-server-browser --env DISPLAY=<YOUR_DISPLAY> -- npx -y @agent-infra/mcp-server-browser browser cdp-endpoint ws-endpoint executable-path output-dir proxy-bypass proxy-server vision

Use cases

  • Automate multi-step browser tasks like booking flights or hotels via natural language instructions
  • Give an AI agent DOM-based or visual (screenshot-driven) control of a web browser
  • Integrate browser automation as one MCP tool alongside other MCP servers in Agent TARS workflows
  • Enable headless or headful agent execution that performs real-world web interactions

io.github.bytedance/mcp-server-browser MCP server FAQ

What does the mcp-server-browser MCP server do?

It provides browser-use access to AI agents over the Model Context Protocol, enabling control of a web browser (via DOM interaction and/or visual/GUI grounding) as part of ByteDance's Agent TARS and UI-TARS Desktop agent stack.

Is mcp-server-browser free to use?

Yes, the code is open source under the Apache License 2.0 and available as the npm package @agent-infra/mcp-server-browser.

How do I install it?

Install via npm with `npm install @agent-infra/mcp-server-browser` or run it through the Agent TARS CLI (`npx @agent-tars/cli@latest`), then mount it as an MCP server so a client like Claude or Cursor can connect to it.

Does it require authentication?

The README does not describe any specific authentication requirement for the browser server itself, though the broader Agent TARS CLI requires an API key for the chosen LLM provider (e.g., --apiKey your-api-key).

What specific tools/functions does this MCP server expose?

The README does not enumerate specific tool names or function signatures for mcp-server-browser; it only describes general browser-use/control capability as part of the hybrid GUI/DOM browser agent.

Can this run headless or with a UI?

The broader Agent TARS project it belongs to supports both headful Web UI and headless server execution modes via its CLI.

README (reference)

Source of truth, from the repository.

<picture> <img alt="Agent TARS Banner" src="./images/tars.png"> </picture> <br/>

Introduction

English | 简体中文

<b>TARS<sup>*</sup></b> is a Multimodal AI Agent stack, currently shipping two projects: Agent TARS and UI-TARS-desktop:

<table> <thead> <tr> <th width="50%" align="center"><a href="#agent-tars">Agent TARS</a></th> <th width="50%" align="center"><a href="#ui-tars-desktop">UI-TARS-desktop</a></th> </tr> </thead> <tbody> <tr> <td align="center"> <video src="https://github.com/user-attachments/assets/c9489936-afdc-4d12-adda-d4b90d2a869d" width="50%"></video> </td> <td align="center"> <video src="https://github.com/user-attachments/assets/e0914ce9-ad33-494b-bdec-0c25c1b01a27" width="50%"></video> </td> </tr> <tr> <td align="left"> <b>Agent TARS</b> is a general multimodal AI Agent stack, it brings the power of GUI Agent and Vision into your terminal, computer, browser and product. <br> <br> It primarily ships with a <a href="https://agent-tars.com/guide/basic/cli.html" target="_blank">CLI</a> and <a href="https://agent-tars.com/guide/basic/web-ui.html" target="_blank">Web UI</a> for usage. It aims to provide a workflow that is closer to human-like task completion through cutting-edge multimodal LLMs and seamless integration with various real-world <a href="https://agent-tars.com/guide/basic/mcp.html" target="_blank">MCP</a> tools. </td> <td align="left"> <b>UI-TARS Desktop</b> is a desktop application that provides a native GUI Agent based on the <a href="https://github.com/bytedance/UI-TARS" target="_blank">UI-TARS</a> model. <br> <br> It primarily ships a <a href="https://github.com/bytedance/UI-TARS-desktop/blob/main/docs/quick-start.md#get-model-and-run-local-operator" target="_blank">local</a> and <a href="https://github.com/bytedance/UI-TARS-desktop/blob/main/docs/quick-start.md#run-remote-operator" target="_blank">remote</a> computer as well as browser operators. </td> </tr> </tbody> </table>

Table of Contents

<!-- START doctoc generated TOC please keep comment here to allow auto update --> <!-- DON'T EDIT THIS SECTION, INSTEAD RE-RUN doctoc TO UPDATE --> <!-- END doctoc generated TOC please keep comment here to allow auto update -->

News

  • [2025-11-05] 🎉 We're excited to announce the release of Agent TARS CLI v0.3.0! This version brings streaming support for multiple tools (shell commands, multi-file structured display), runtime settings with timing statistics for tool calls and deep thinking, Event Stream Viewer for data flow tracking and debugging. Additionally, it features exclusive support for AIO agent Sandbox as isolated all-in-one tools execution environment.
  • [2025-06-25] We released an Agent TARS Beta and Agent TARS CLI - Introducing Agent TARS Beta, a multimodal AI agent that aims to explore a work form that is closer to human-like task completion through rich multimodal capabilities (such as GUI Agent, Vision) and seamless integration with various real-world tools.
  • [2025-06-12] - 🎁 We are thrilled to announce the release of UI-TARS Desktop v0.2.0! This update introduces two powerful new features: Remote Computer Operator and Remote Browser Operator—both completely free. No configuration required: simply click to remotely control any computer or browser, and experience a new level of convenience and intelligence.
  • [2025-04-17] - 🎉 We're thrilled to announce the release of new UI-TARS Desktop application v0.1.0, featuring a redesigned Agent UI. The application enhances the computer using experience, introduces new browser operation features, and supports the advanced UI-TARS-1.5 model for improved performance and precise control.
  • [2025-02-20] - 📦 Introduced UI TARS SDK, is a powerful cross-platform toolkit for building GUI automation agents.
  • [2025-01-23] - 🚀 We updated the Cloud Deployment section in the 中文版: GUI模型部署教程 with new information related to the ModelScope platform. You can now use the ModelScope platform for deployment.
<br>

Agent TARS

<p> <a href="https://npmjs.com/package/@agent-tars/cli?activeTab=readme"><img src="https://img.shields.io/npm/v/@agent-tars/cli?style=for-the-badge&colorA=1a1a2e&colorB=3B82F6&logo=npm&logoColor=white" alt="npm version" /></a> <a href="https://npmcharts.com/compare/@agent-tars/cli?minimal=true"><img src="https://img.shields.io/npm/dm/@agent-tars/cli.svg?style=for-the-badge&colorA=1a1a2e&colorB=0EA5E9&logo=npm&logoColor=white" alt="downloads" /></a> <a href="https://nodejs.org/en/about/previous-releases"><img src="https://img.shields.io/node/v/@agent-tars/cli.svg?style=for-the-badge&colorA=1a1a2e&colorB=06B6D4&logo=node.js&logoColor=white" alt="node version"></a> <a href="https://discord.gg/HnKcSBgTVx"><img src="https://img.shields.io/badge/Discord-Join%20Community-5865F2?style=for-the-badge&logo=discord&logoColor=white" alt="Discord Community" /></a> <a href="https://twitter.com/agent_tars"><img src="https://img.shields.io/badge/Twitter-Follow%20%40agent__tars-1DA1F2?style=for-the-badge&logo=twitter&logoColor=white" alt="Official Twitter" /></a> <a href="https://applink.larkoffice.com/client/chat/chatter/add_by_link?link_token=deen76f4-ea3c-4964-93a3-78f126f39651"><img src="https://img.shields.io/badge/飞书群-加入交流群-00D4AA?style=for-the-badge&logo=lark&logoColor=white" alt="飞书交流群" /></a> <a href="https://deepwiki.com/bytedance/UI-TARS-desktop"><img src="https://img.shields.io/badge/DeepWiki-Ask%20AI-8B5CF6?style=for-the-badge&logo=gitbook&logoColor=white" alt="Ask DeepWiki" /></a> </p>

<b>Agent TARS</b> is a general multimodal AI Agent stack, it brings the power of GUI Agent and Vision into your terminal, computer, browser and product. <br> <br> It primarily ships with a <a href="https://agent-tars.com/guide/basic/cli.html" target="_blank">CLI</a> and <a href="https://agent-tars.com/guide/basic/web-ui.html" target="_blank">Web UI</a> for usage. It aims to provide a workflow that is closer to human-like task completion through cutting-edge multimodal LLMs and seamless integration with various real-world <a href="https://agent-tars.com/guide/basic/mcp.html" target="_blank">MCP</a> tools.

Showcase

Please help me book the earliest flight from San Jose to New York on September 1st and the last return flight on September 6th on Priceline

https://github.com/user-attachments/assets/772b0eef-aef7-4ab9-8cb0-9611820539d8

<br> <table> <thead> <tr> <th width="50%" align="center">Booking Hotel</th> <th width="50%" align="center">Generate Chart with extra MCP Servers</th> </tr> </thead> <tbody> <tr> <td align="center"> <video src="https://github.com/user-attachments/assets/c9489936-afdc-4d12-adda-d4b90d2a869d" width="50%"></video> </td> <td align="center"> <video src="https://github.com/user-attachments/assets/a9fd72d0-01bb-4233-aa27-ca95194bbce9" width="50%"></video> </td> </tr> <tr> <td align="left"> <b>Instruction:</b> <i>I am in Los Angeles from September 1st to September 6th, with a budget of $5,000. Please help me book a Ritz-Carlton hotel closest to the airport on booking.com and compile a transportation guide for me</i> </td> <td align="left"> <b>Instruction:</b> <i>Draw me a chart of Hangzhou's weather for one month</i> </td> </tr> </tbody> </table>

For more use cases, please check out #842.

Core Features

  • 🖱️ One-Click Out-of-the-box CLI - Supports both headful Web UI and headless server execution.
  • 🌐 Hybrid Browser Agent - Control browsers using GUI Agent, DOM, or a hybrid strategy.
  • 🔄 Event Stream - Protocol-driven Event Stream drives Context Engineering and Agent UI.
  • 🧰 MCP Integration - The kernel is built on MCP and also supports mounting MCP Servers to connect to real-world tools.

Quick Start

<img alt="Agent TARS CLI" src="https://agent-tars.com/agent-tars-cli.png"> 
# Launch with `npx`.
npx @agent-tars/cli@latest

# Install globally, required Node.js >= 22
npm install @agent-tars/cli@latest -g

# Run with your preferred model provider
agent-tars --provider volcengine --model doubao-1-5-thinking-vision-pro-250428 --apiKey your-api-key
agent-tars --provider anthropic --model claude-3-7-sonnet-latest --apiKey your-api-key

Visit the comprehensive Quick Start guide for detailed setup instructions.

Documentation

🌟 Explore Agent TARS Universe 🌟

<table> <thead> <tr> <th width="20%" align="center">Category</th> <th width="30%" align="center">Resource Link</th> <th width="50%" align="left">Description</th> </tr> </thead> <tbody> <tr> <td align="center">🏠 <strong>Central Hub</strong></td> <td align="center"> <a href="https://agent-tars.com"> <img src="https://img.shields.io/badge/Visit-Website-4F46E5?style=for-the-badge&logo=globe&logoColor=white" alt="Website" /> </a> </td> <td align="left">Your gateway to Agent TARS ecosystem</td> </tr> <tr> <td align="center">📚 <strong>Quick Start</strong></td> <td align="center"> <a href="https://agent-tars.com/guide/get-started/quick-start.html"> <img src="https://img.shields.io/badge/Get-Started-06B6D4?style=for-the-badge&logo=rocket&logoColor=white" alt="Quick Start" /> </a> </td> <td align="left">Zero to hero in 5 minutes</td> </tr> <tr> <td align="center">🚀 <strong>What's New</strong></td> <td align="center"> <a href="https://agent-tars.com/beta"> <img src="https://img.shields.io/badge/Read-Blog-F59E0B?style=for-the-badge&logo=rss&logoColor=white" alt="Blog" /> </a> </td> <td align="left">Discover cutting-edge features & vision</td> </tr> <tr> <td align="center">🛠️ <strong>Developer Zone</strong></td> <td align="center"> <a href="https://agent-tars.com/guide/get-started/introduction.html"> <img src="https://img.shields.io/badge/View-Docs-10B981?style=for-the-badge&logo=gitbook&logoColor=white" alt="Docs" /> </a> </td> <td align="left">Master every command & features</td> </tr> <tr> <td align="center">🎯 <strong>Showcase</strong></td> <td align="center"> <a href="https://github.com/bytedance/UI-TARS-desktop/issues/842"> <img src="https://img.shields.io/badge/View-Examples-8B5CF6?style=for-the-badge&logo=github&logoColor=white" alt="Examples" /> </a> </td> <td align="left">View use cases built by the official and community</td> </tr> <tr> <td align="center">🔧 <strong>Reference</strong></td> <td align="center"> <a href="https://agent-tars.com/api/"> <img src="https://img.shields.io/badge/API-Reference-EF4444?style=for-the-badge&logo=book&logoColor=white" alt="API" /> </a> </td> <td align="left">Complete technical reference</td> </tr> </tbody> </table> <br/> <br/> <br/>

UI-TARS Desktop

<p align="center"> <img alt="UI-TARS" width="260" src="./apps/ui-tars/resources/icon.png"> </p>

UI-TARS Desktop is a native GUI agent for your local computer, driven by UI-TARS and Seed-1.5-VL/1.6 series models.

<div align="center"> <p> &nbsp&nbsp 📑 <a href="https://arxiv.org/abs/2501.12326">Paper</a> &nbsp&nbsp | 🤗 <a href="https://huggingface.co/ByteDance-Seed/UI-TARS-1.5-7B">Hugging Face Models</a>&nbsp&nbsp | &nbsp&nbsp🫨 <a href="https://discord.gg/pTXwYVjfcs">Discord</a>&nbsp&nbsp | &nbsp&nbsp🤖 <a href="https://www.modelscope.cn/collections/UI-TARS-bccb56fa1ef640">ModelScope</a>&nbsp&nbsp <br> 🖥️ Desktop Application &nbsp&nbsp | &nbsp&nbsp 👓 <a href="https://github.com/web-infra-dev/midscene">Midscene (use in browser)</a> &nbsp&nbsp </p> </div>

Showcase

<!-- // FIXME: Choose only two demo, one local computer and one remote computer showcase. -->
InstructionLocal OperatorRemote Operator
Please help me open the autosave feature of VS Code and delay AutoSave operations for 500 milliseconds in the VS Code setting.<video src="https://github.com/user-attachments/assets/e0914ce9-ad33-494b-bdec-0c25c1b01a27" height="300" /><video src="https://github.com/user-attachments/assets/01e49b69-7070-46c8-b3e3-2aaaaec71800" height="300" />
Could you help me check the latest open issue of the UI-TARS-Desktop project on GitHub?<video src="https://github.com/user-attachments/assets/3d159f54-d24a-4268-96c0-e149607e9199" height="300" /><video src="https://github.com/user-attachments/assets/072fb72d-7394-4bfa-95f5-4736e29f7e58" height="300" />

Features

  • 🤖 Natural language control powered by Vision-Language Model
  • 🖥️ Screenshot and visual recognition support
  • 🎯 Precise mouse and keyboard control
  • 💻 Cross-platform support (Windows/MacOS/Browser)
  • 🔄 Real-time feedback and status display
  • 🔐 Private and secure - fully local processing

Quick Start

See Quick Start

Contributing

See CONTRIBUTING.md.

License

This project is licensed under the Apache License 2.0.

Citation

If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil:

@article{qin2025ui,
  title={UI-TARS: Pioneering Automated GUI Interaction with Native Agents},
  author={Qin, Yujia and Ye, Yining and Fang, Junjie and Wang, Haoming and Liang, Shihao and Tian, Shizuo and Zhang, Junda and Li, Jiahao and Li, Yunxin and Huang, Shijue and others},
  journal={arXiv preprint arXiv:2501.12326},
  year={2025}
}

Related MCP servers

io.github.bytedance/mcp-server-commands

Active

Run arbitrary shell commands from an MCP-connected AI agent.

37k
TypeScript
Apache-2.0
View repository →

io.github.bytedance/mcp-server-filesystem

Active

Filesystem access MCP server from ByteDance's UI-TARS/Agent TARS ecosystem.

37k
TypeScript
Apache-2.0
View repository →

io.github.bytedance/mcp-server-search

Active

Adds web search capability to MCP-enabled AI agents like Agent TARS.

37k
TypeScript
Apache-2.0
View repository →