# Browser Use Cloud SDK

> The most SOTA browser agent and the most scalable browser infrastructure. Built on the largest AI browser automation open-source library on GitHub with almost 100k stars. Web scraping, data extraction, form filling, and headless browser automation via natural language. Anti-detection, CAPTCHA solving, residential proxies. Alternative to Playwright, Puppeteer, and Selenium for AI-powered automation.

- GitHub: https://github.com/browser-use/browser-use
- Dashboard: https://cloud.browser-use.com
- API key: https://cloud.browser-use.com/settings?tab=api-keys&new=1
- Open Source: https://github.com/browser-use/browser-use
- Docs: https://docs.browser-use.com
- OpenAPI spec (v3): https://docs.browser-use.com/cloud/openapi/v3.json
- Chat UI example: https://docs.browser-use.com/cloud/tutorials/chat-ui — Full end-to-end example with live browser, streaming, auth. Best starting point to build a prototype.
- Stealth benchmark: https://browser-use.com/posts/stealth-benchmark
- Agent benchmark (online Mind2Web): https://browser-use.com/posts/online-mind2web-benchmark
- Blog: https://browser-use.com/posts

Install (check if already installed and update to latest):
- Python: `pip install --upgrade browser-use-sdk`
- TypeScript: `npm install browser-use-sdk@latest`

Set API key (starts with `bu_`):
```
export BROWSER_USE_API_KEY=bu_your_key_here
```


## Get Started
- [Quick start](https://docs.browser-use.com/cloud/quickstart): State-of-the-art AI browser automation with stealth browsers, CAPTCHA solving, residential proxies, and managed infrastructure.
- [Prompt for Vibecoders](https://docs.browser-use.com/cloud/vibecoding): Complete Cloud SDK reference for AI coding agents.


## Agent
- [Introduction](https://docs.browser-use.com/cloud/agent/quickstart): Easiest way to automate the web. Tell this agent in natural language what it should do, and it can interact with the web like a human.
- [Structured output](https://docs.browser-use.com/cloud/agent/structured-output): Get validated, typed data back from agent tasks.
- [Follow-up tasks](https://docs.browser-use.com/cloud/agent/follow-up-tasks): Run multiple tasks in the same browser session.
- [Live messages](https://docs.browser-use.com/cloud/agent/streaming): Poll the agent's message history in real time to build custom UIs or monitor progress.
- [Workspaces & files](https://docs.browser-use.com/cloud/agent/workspaces): Persistent file storage across sessions. Upload input files, download results, and share data across tasks.
- [Human in the loop](https://docs.browser-use.com/cloud/agent/human-in-the-loop): Let a human interact with the live browser while the agent is running. Useful for approvals, payments, complex auth flows, or reviewing agent work before continuing.

## Browser
- [Introduction Stealth](https://docs.browser-use.com/cloud/browser/stealth): Best stealth on the planet. We fork Chromium to give agents access to all websites.
- [Proxies](https://docs.browser-use.com/cloud/browser/proxies): Residential proxies in 195+ countries. On by default.
- [Live preview & recording](https://docs.browser-use.com/cloud/browser/live-preview): Watch the agent's browser in real time. Embed it in your app.

## Authentication
- [Profiles](https://docs.browser-use.com/cloud/guides/authentication): Persistent browser state — cookies, localStorage, saved passwords. Login once, reuse across sessions.
- [Sync local and cloud cookies](https://docs.browser-use.com/cloud/guides/profile-sync): Sync your local browser cookies to the cloud — instantly authenticate without managing credentials.
- [Playwright, Puppeteer, Selenium](https://docs.browser-use.com/cloud/browser/playwright-puppeteer-selenium): Connect your automation framework to Browser Use's stealth infrastructure via CDP.


## Integrations
- [OpenClaw](https://docs.browser-use.com/cloud/tutorials/integrations/openclaw): Give OpenClaw agents browser automation with Browser Use — via CDP or the CLI skill.
- [MCP Server](https://docs.browser-use.com/cloud/guides/mcp-server): Run browser automation tasks from your AI coding assistant. Connect to Claude, Cursor, Windsurf, or any MCP client.
- [Webhooks](https://docs.browser-use.com/cloud/guides/webhooks): Receive real-time notifications when tasks complete. Configure webhook endpoints for async task monitoring.
- [n8n](https://docs.browser-use.com/cloud/tutorials/integrations/n8n): Use Browser Use as an HTTP node in n8n workflows.

## Tutorials
- [Chat UI](https://docs.browser-use.com/cloud/tutorials/chat-ui): Full end-to-end example. Build a chat UI with live browser preview, authentication, streaming messages, and session management. Best starting point for integrating Browser Use as a main agent or sub-agent into your app.
- [FAQ](https://docs.browser-use.com/cloud/faq): Common questions and solutions.

## Legacy (v2)
- [Agent (v2)](https://docs.browser-use.com/cloud/legacy/agent): V2 agent models and file handling.
- [Skills](https://docs.browser-use.com/cloud/legacy/skills): Turn any website into a deterministic API endpoint. Create once, call repeatedly.
- [1Password & 2FA](https://docs.browser-use.com/cloud/guides/1password): Auto-fill passwords and TOTP codes from 1Password during agent tasks.
- [Secrets](https://docs.browser-use.com/cloud/guides/secrets): Pass domain-scoped credentials to the agent securely.