RELATED 01
Browser Use
OFFICIAL SERVERLocal MCP server that drives a Chrome or Chromium browser step by step, with an agent fallback
120kGitHub stars
8.1M/moPyPI downloads
119popularity score
Browser Use MCP Server
Lets an assistant drive a browser with Puppeteer: navigate, click, fill forms, read pages, take screenshots.
By Bytedance Inc. TypeScript Local · stdio Browser automation
56 popularity score ★ 40k ↓ 792/wk
Connect your client
Local (npm)
Add in one click
Opens the app, which asks before it adds the server.
Or add it by command or config
Claude CodeTerminal
claude mcp add browser-use -- npx -y @agent-infra/mcp-server-browserClaude Desktopclaude_desktop_config.json
{
"mcpServers": {
"browser-use": {
"command": "npx",
"args": [
"-y",
"@agent-infra/mcp-server-browser"
]
}
}
}Cursor~/.cursor/mcp.json
{
"mcpServers": {
"browser-use": {
"command": "npx",
"args": [
"-y",
"@agent-infra/mcp-server-browser"
]
}
}
}VS Code.vscode/mcp.json
{
"servers": {
"browser-use": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@agent-infra/mcp-server-browser"
]
}
}
}Codex~/.codex/config.toml
[mcp_servers.browser-use]
command = "npx"
args = ["-y", "@agent-infra/mcp-server-browser"]Gemini CLITerminal
gemini mcp add --scope user browser-use npx -- -y @agent-infra/mcp-server-browserWhat it does
The Browser Use MCP Server (@agent-infra/mcp-server-browser) is ByteDance's browser automation server from the UI-TARS Desktop project. It drives a browser through Puppeteer and gives the model structured data from the page: indexed lists of clickable elements, page text, markdown and links, so it can act without a vision model. An optional vision mode adds screenshot capture and coordinate clicks for layouts where structured data is not enough.
Its tools navigate, go back and forward, click, hover, fill inputs, select options, press keys, scroll, run JavaScript, manage tabs, list downloads and take screenshots; console logs, downloads and screenshots are also offered as resources.
It runs locally with npx @agent-infra/mcp-server-browser@latest (Node.js 18 or newer), using Chrome, Edge or Firefox, headed by default or headless, or connects to a remote browser through a CDP or WebSocket endpoint. With --port it serves SSE and Streamable HTTP for remote use, where viewport size, vision coordinate factors and user agent can be set per request with HTTP headers. This page covers the MCP server package; UI-TARS Desktop itself is a separate desktop agent app in the same repository.
Tools 23
browser_click- Click an element on the page
browser_close- Close the browser when the task is done and the browser is not needed anymore
browser_close_tab- Close the current tab
browser_evaluate- Execute JavaScript in the browser console
browser_form_input_fill- Fill out an input field
browser_get_clickable_elements- Get the clickable or hoverable or selectable elements on the current page, don't call this tool multiple times
browser_get_download_list- Get the list of downloaded files
browser_get_markdown- Get the markdown content of the current page
browser_get_text- Get the text content of the current page
browser_go_back- Go back to the previous page
browser_go_forward- Go forward to the next page
browser_hover- Hover an element on the page
browser_navigate- Navigate to a URL
browser_new_tab- Open a new tab
browser_press_key- Press a key on the keyboard
browser_read_links- Get all links on the current page
browser_screenshot- Take a screenshot of the current page or a specific element
browser_scroll- Scroll the page
browser_select- Select an element on the page with index
browser_switch_tab- Switch to a specific tab
browser_tab_list- Get the list of tabs
browser_vision_screen_capture- Take a screenshot of the current page for vision mode
browser_vision_screen_click- Click left mouse button on the page with vision and snapshot