</>MCP Agents Market
Skill

browser-use

by browser-use109.4kPythonUpdated 2026-08-16

🌐 Make websites accessible for AI agents. Automate tasks online with ease.

Claude CodeCodexCursorHermesOpenClaw

Browser Use is an agent skill that enables AI agents to control web browsers like humans do—opening pages, clicking elements, typing text, and filling forms to automate online tasks. It works as both a Python library for building custom web automation and as an MCP skill that integrates with AI coding assistants. Developers can use it to automate data extraction, form submissions, account management, and other repetitive browser-based workflows with any major LLM provider.

Key Features

Programmatic browser control: AI agents can navigate, click, type, and extract data from web pages autonomously
Python library (3.11+) with async support for building custom automation scripts at scale
MCP skill integration for Claude Code, Codex, Cursor, and other AI coding assistants
Multi-LLM support including OpenAI, Anthropic Claude, Google Gemini, and optimized ChatBrowserUse models
Custom tool support to extend agent capabilities with domain-specific functions
Authentication handling through real browser profiles and session persistence
Cloud API option for stealth browsers, proxy rotation, CAPTCHA solving, and production scalability
Open-source MIT license with optional fully-hosted cloud service for complex tasks

Use Cases

  • 01Automating job application form filling with resume data
  • 02Extracting structured data from social media profiles and exporting to CSV
  • 03Monitoring websites and scraping content on schedules or in parallel
  • 04Running QA tests across web applications with browser interaction
  • 05Managing accounts and performing routine tasks like video uploads or content posting
  • 06Building products that embed intelligent browser automation capabilities

Related Skills

View more

browser-use — FAQ

What is browser-use and what does it do?+

Browser-use is an agent skill that lets AI agents interact with websites by controlling a real web browser—clicking, typing, navigating, and extracting data just like a human would. It automates online tasks based on natural language descriptions.

How do I install and connect browser-use to my AI agent?+

For AI coding assistants like Claude Code or Cursor, paste the setup prompt provided in the quickstart to auto-install. For Python projects, run 'uv add browser-use' or 'pip install browser-use' (Python 3.11+), add your LLM API key to a .env file, and create an Agent instance in your code.

Which AI clients work with browser-use?+

Browser-use works as an MCP skill with Claude Code, Codex, Cursor, Hermes, OpenClaw, and similar AI coding assistants. As a Python library, it integrates with any LangChain-compatible client supporting OpenAI, Anthropic, Google, or ChatBrowserUse models.

Do I need API keys to use browser-use?+

Yes, you need an API key from an LLM provider. You can use Browser Use Cloud (BROWSER_USE_API_KEY) which accesses multiple models, or bring your own keys from OpenAI, Anthropic, or Google. Local models via Ollama are also supported.

Is browser-use free to use?+

Yes, the open-source library is free under the MIT license. You only pay for the LLM provider API calls. Browser Use also offers a paid cloud service with advanced features like stealth browsers, CAPTCHA solving, and scalable infrastructure.

What's the best model to use with browser-use?+

ChatBrowserUse models are optimized specifically for browser automation and complete tasks 3-5x faster with state-of-the-art accuracy. You can also use Claude Sonnet/Opus, GPT-4/5, Gemini, or local models depending on your needs.

How do I install browser-use?+

Open the source repository on GitHub and follow its README. browser-use is a skill — MCP Agents Market links you directly to the official repo.

Is browser-use free?+

browser-use is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches