Browser Use

Browser automation platform that lets AI agents perform web tasks in natural language. Agents extract data, automate workflows, run tests, and monitor pages without writing selectors or scripts, with optional stealth browsers and proxies for production use.

Browser Use differentiates by coupling LLMs directly to browser control: you describe goals in plain language and the agent navigates, clicks, types, and extracts. It offers Skill APIs to turn any website into a reliable API endpoint, create once and call forever. Supports custom models purpose-built for browser automation, plus anti-detect browsers, CAPTCHA solving, and global proxies when running on their cloud.

Key capabilities:

• Natural language web agents: extract, automate, test, and monitor via task descriptions
• Stealth browsers with anti-detect, CAPTCHA solving, and 195+ country proxies, zero config
• Skill APIs to expose websites as API endpoints for repeated automation
• Python SDK with uv, compatible with Google, Anthropic, or Browser Use hosted models
• CLI for direct control: open URLs, inspect state, click, type, screenshot, close
• Claude skills integration and custom tools via action decorators
• Local or cloud execution with optional managed infrastructure

Use cases: AI agents that need web access for research, data extraction, or task execution, workflow automation for e-commerce and admin operations, and turning complex sites into callable APIs. Developers integrate via the Python SDK, use the CLI for scripting, or add the Claude skill for AI-assisted browsing.