{ "markdown": "\n\n \n \n \"Shows\n\n\n
\n \n \n \n \"The\n \n
\n\n
\n\"Browser-Use\n
\n\n---\n\n
\n\"Demos\"\n\"\"\n\"Docs\"\n\"\"\n\"Blog\"\n\"\"\n\"Merch\"\n\"\"\n\"Github\n\"\"\n\"Twitter\"\n\"\"\n\"Discord\"\n\"\"\n\"Browser-Use\n
\n\n
\n\n# What can Browser Use do?\n\nBrowser Use lets an AI agent use a web browser the same way humans do — it opens pages, clicks buttons, types, and fills in forms. You describe the task, and it completes it. For example, you can have it:\n\n\n### 📋 Fill Forms\n#### Task: \"Fill in this job application with my resume and information.\"\n\n![Job Application Demo](https://github.com/user-attachments/assets/57611d8e-0474-4de6-84b7-37a0c0cd27e7)\n\n[Example code ↗](https://github.com/browser-use/browser-use/blob/main/examples/use-cases/apply_to_job.py)\n\n\n### 🍎 Extract data\n#### Task: \"Extract structured data about my followers and export it as a CSV.\"\n\nhttps://github.com/user-attachments/assets/485fd3ec-61b9-4afc-9e86-ee9b85acb592\n\n[Browser Use Cloud Docs ↗](https://docs.browser-use.com/cloud/quickstart)\n\n\n
\n\n# Start with Browser Use Cloud\n\nEligible new Google, GitHub or Microsoft signups get a one-time **$15 Cloud credit**. No card required; email/password signups do not qualify. [Pricing and eligibility](https://browser-use.com/pricing.md).\n\n- **Hosted agent:** [V4 quickstart](https://docs.browser-use.com/cloud/agent/quickstart), using `browser-use-sdk`.\n- **Your own agent:** [Managed browser quickstart](https://docs.browser-use.com/cloud/browser/quickstart).\n- **Local framework:** the `browser-use` Python library below.\n\n# Quickstart\n\nIf you want to use Browser Use in your agent (Claude Code, Codex, Cursor, Hermes, OpenClaw, etc.), paste this prompt, and it sets everything up itself:\n\n```text\nInstall or upgrade browser-use to the latest stable version with uv using Python 3.12, run `browser-use skill install` to register the skill, and connect it to my browser. If setup or connection fails, follow https://github.com/browser-use/browser-harness/blob/main/install.md.\n```\n\nThen tell your agent what you want done.\n\n
\n\n# Python library: the easiest way to automate the web\n\nWant to automate the web at scale, from your own code, and with any LLM? Use the Python library:\n\n**1. Install Browser Use (Python >= 3.11):**\n\n```bash\nuv add browser-use\n# or: pip install browser-use\n```\n\n**2. Add your LLM API key to `.env`**. Get one from [Browser Use Cloud](https://cloud.browser-use.com/new-api-key?utm_source=github&utm_medium=readme-quickstart-api-key), or bring your own provider key:\n\n```bash\n# .env\nBROWSER_USE_API_KEY=your-key\n# GOOGLE_API_KEY=your-key\n# ANTHROPIC_API_KEY=your-key\n```\n\n**3. Run your first agent:**\n\n```python\nimport asyncio\n\nfrom browser_use import Agent, ChatBrowserUse\n\nasync def main():\n agent = Agent(\n task=\"Find the number of stars of the browser-use repo\",\n llm=ChatBrowserUse(model='openai/gpt-5.5'),\n # llm=ChatBrowserUse(model='bu-2-0-mini-preview'), # Browser Use's optimized model\n # llm=ChatOpenAI(model='gpt-5.5'),\n # llm=ChatAnthropic(model='claude-opus-4-8'), # Sonnet also works well\n )\n history = await agent.run()\n\nif __name__ == \"__main__\":\n asyncio.run(main())\n```\n\nCheck out the [library docs](https://docs.browser-use.com/open-source/introduction) and the [cloud docs](https://docs.cloud.browser-use.com?utm_source=github&utm_medium=readme-cloud-docs) for more!\n\n
\n\n# Open Source vs Cloud\n\n\n \n \n \"BU\n\n\nWe benchmark Browser Use across 100 real-world browser tasks. Full benchmark is open source: **[browser-use/benchmark](https://github.com/browser-use/benchmark)**.\n\nBrowser Use is also **#1 on the [Odysseys leaderboard](https://odysseysbench.com/leaderboard)** with an 87.4% average, ahead of computer-use agents from OpenAI, Anthropic, Google, and Microsoft. Odysseys measures the agent's performance on 200 long-horizon web tasks.\n\n**Use the Open-Source Agent**\n- Free, and runs on your own machine\n- Deep code-level integration and control: pick your LLM, customize the agent's behavior\n- We recommend pairing it with our [cloud browsers](https://docs.browser-use.com/open-source/customize/browser/remote) for leading stealth, proxy rotation, and scaling\n\n**Use the [Fully-Hosted Cloud Agent](https://cloud.browser-use.com?utm_source=github&utm_medium=readme-hosted-agent) (recommended)**\n- Much more powerful agent for complex tasks (see plot above)\n- Easiest way to start and scale\n- Best stealth with proxy rotation and captcha solving\n- 1000+ integrations (Gmail, Slack, Notion, and more)\n- Persistent filesystem and memory\n- Rerunnable scripts fetch live data, even when sites change ([guide](https://docs.browser-use.com/cloud/agent/scripts))\n\n```sh\ncurl -X POST https://api.browser-use.com/api/v4/runs \\\n -H \"X-Browser-Use-API-Key: $BROWSER_USE_API_KEY\" \\\n -H \"Content-Type: application/json\" \\\n -d '{\"task\": \"Your task\"}'\n```\n\n
\n\n## Integrations, hosting, custom tools, MCP, and more on our [Docs ↗](https://docs.browser-use.com)\n\n
\n\n# FAQ\n\n
\nShould I use the CLI vs. the Python library?\n\n**Use the CLI** if you already have an agent (Claude Code, Codex, Cursor, Hermes, OpenClaw, etc.) that you want to complete browser tasks for you. The agent installs the skill once (see [Quickstart](#quickstart)) and can then control the browser. Examples:\n- \"Upload this video to YouTube\"\n- \"Compare these three laptops and give me a table with prices\"\n- \"Fill in this job application with my resume\"\n\n**Use the Python library** when you are building software that automates the web. Examples:\n- Run many tasks on a schedule or in parallel (scraping, monitoring, QA)\n- Embed a browser agent into your own product\n- Custom tools, custom system prompts, structured output, fine-grained browser control\n\nRule of thumb: one-off tasks through an agent → CLI. Repeatable automation in code → Python library.\n
\n\n
\nWhat's the best model to use?\n\nWe optimized **ChatBrowserUse()** specifically for browser automation tasks. On avg it completes tasks 3-5x faster than other models with SOTA accuracy.\n\nFor pricing and other LLM providers, see our [supported models documentation](https://docs.browser-use.com/supported-models).\n
\n\n
\nCan I use Claude / GPT / Gemini through ChatBrowserUse?\n\nYes. `ChatBrowserUse` accepts provider-prefixed model ids, so a single `BROWSER_USE_API_KEY` reaches all of them — no separate OpenAI/Anthropic/Google keys required:\n\n```python\nfrom browser_use import Agent, ChatBrowserUse\n\nllm = ChatBrowserUse(model='anthropic/claude-sonnet-4-6') # or 'openai/gpt-5.5', 'google/gemini-3-pro'\nagent = Agent(task='...', llm=llm)\n```\n\nFor the best speed and cost we still recommend the default `bu-*` models.\n
\n\n
\nShould I use the Browser Use system prompt with the open-source preview model?\n\nYes. If you use `ChatBrowserUse(model='browser-use/bu-30b-a3b-preview')` with a normal `Agent(...)`, Browser Use still sends its default agent system prompt for you.\n\nYou do **not** need to add a separate custom \"Browser Use system message\" just because you switched to the open-source preview model. Only use `extend_system_message` or `override_system_message` when you intentionally want to customize the default behavior for your task.\n\nIf you want the best default speed/accuracy, we still recommend the newer hosted `bu-*` models. If you want the open-source preview model, the setup stays the same apart from the `model=` value.\n
\n\n
\nCan I use custom tools with the agent?\n\nYes! You can add custom tools to extend the agent's capabilities:\n\n```python\nfrom browser_use import Tools\n\ntools = Tools()\n\n@tools.action(description='Description of what this tool does.')\ndef custom_tool(param: str) -> str:\n return f\"Result: {param}\"\n\nagent = Agent(\n task=\"Your task\",\n llm=llm,\n browser=browser,\n tools=tools,\n)\n```\n\n
\n\n
\nCan I use this for free?\n\nYes! Browser-Use is open source and free to use. You only need to choose an LLM provider (like OpenAI, Google, ChatBrowserUse, or run local models with Ollama).\n
\n\n
\nTerms of Service\n\nThis open-source library is licensed under the MIT License. For Browser Use services & data policy, see our [Terms of Service](https://browser-use.com/legal/terms-of-service) and [Privacy Policy](https://browser-use.com/privacy/).\n
\n\n
\nHow do I handle authentication?\n\nCheck out our authentication examples:\n- [Using real browser profiles](https://github.com/browser-use/browser-use/blob/main/examples/browser/real_browser.py) - Reuse your existing Chrome profile with saved logins\n- If you want to use temporary accounts with inbox, choose AgentMail\n- To sync your auth profile with a remote browser, install `profile-use` for your platform from the [official releases](https://github.com/browser-use/profile-use-releases/releases/latest), then follow the [profile sync guide](https://github.com/browser-use/browser-harness/blob/main/interaction-skills/profile-sync.md).\n\nThese examples show how to maintain sessions and handle authentication seamlessly.\n
\n\n
\nHow do I solve CAPTCHAs?\n\nFor CAPTCHA handling, you need better browser fingerprinting and proxies. Use [Browser Use Cloud](https://cloud.browser-use.com?utm_source=github&utm_medium=readme-faq-captcha) which provides stealth browsers designed to avoid detection and CAPTCHA challenges.\n
\n\n
\nHow do I go into production?\n\nChrome can consume a lot of memory, and running many agents in parallel can be tricky to manage.\n\nFor production use cases, use our [Browser Use Cloud API](https://cloud.browser-use.com?utm_source=github&utm_medium=readme-faq-production) which handles:\n- Scalable browser infrastructure\n- Memory management\n- Proxy rotation\n- Stealth browser fingerprinting\n- High-performance parallel execution\n
\n\n
\n\n## Citation\n\nIf you use Browser Use in your research or project, please cite:\n\n```bibtex\n@software{browser_use2024,\n author = {Müller, Magnus and Žunič, Gregor},\n title = {Browser Use: Enable AI to control your browser},\n year = {2024},\n publisher = {GitHub},\n url = {https://github.com/browser-use/browser-use}\n}\n```\n\n
\n\n
\n\n**Tell your computer what to do, and it gets it done.**\n\n\n\n[![Twitter Follow](https://img.shields.io/twitter/follow/Magnus?style=social)](https://x.com/intent/user?screen_name=mamagnus00)\n   \n[![Twitter Follow](https://img.shields.io/twitter/follow/Gregor?style=social)](https://x.com/intent/user?screen_name=gregpr07)\n\n
\n\n
Made with ❤️ in Zurich and San Francisco
\n", "bytes": 13627, "sha": "d5c8add03c5483b8136b1282187421f33a1b7d8eebad6937379a692e4d3fa254", "repo_slug": "browser-use/browser-use", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/skl_browser_use_browser_use_browser_use_164ea532/readme" }