← Back to Discover

browser-use

Open-source Python library for building AI agents that drive a web browser, with an optional paid cloud service for stealth browsers and a hosted agent API.

Browser automationAgent frameworks
Visit website ↗

What it does

Browser Use is an open-source browser agent written for Python, with a TypeScript version also mentioned. You give an agent a task in plain language and it opens a browser, works through the pages and returns a result. It can use a local browser, a Browser Use Cloud browser, or an existing remote browser over CDP. The project offers three routes. The hosted cloud runs both the agent and the browser through an API. A CLI gives an existing coding agent access to your browser. The Python library runs the agent inside your own code, with your choice of model, custom tools and structured output. The README recommends its own model, BU2, but also supports providers such as OpenAI, Anthropic and Google, and a local model through Ollama. The library is free and MIT-licensed. Model inference and hosted browsers are billed separately: the README lists cloud browsers at $0.02 per browser-hour plus network traffic, with stealth, residential proxies, CAPTCHA solving and a live view URL. It also notes that no setup guarantees every CAPTCHA can be avoided or solved.

Capabilities

  • Natural-language browser tasks run by an AI agent
  • Use a local browser, a cloud browser or a remote browser over CDP
  • Custom tools, structured output and editable system prompt
  • Works with several model providers, plus local models through Ollama
  • CLI that gives an existing coding agent browser access
  • Hosted agent API and cloud browsers with stealth and proxies

When to use it

  • You are building a web agent in Python that has to fill forms or navigate several pages
  • You want your coding agent to control a real browser from the command line
  • You need managed browsers for scripts running in Docker, CI or serverless setups