Browser-Use MCP

How to Install and Use Browser-Use MCP

If you want to automate browser tasks easily, Browser-Use MCP is a great tool to try. It helps you control a browser using AI to do jobs like filling forms, shopping, or looking up information. Here’s a simple guide to get Browser-Use MCP up and running and see it in action.

Setting Up Your Environment

Before using Browser-Use MCP, you need to prepare your computer. This is the first step to make sure everything runs smoothly.

Start by creating your environment with a tool called “uv” which works with Python 3.11 or higher. Open your terminal or command line and run:

uv init

This command gets everything ready for installing Browser-Use MCP and other tools you’ll need.

Installing Browser-Use MCP

Once your environment is ready, you can install the Browser-Use package.

Run these two commands in your terminal. They make sure you have the latest version because Browser-Use updates every day:

uv add browser-use
uv sync

Now Browser-Use MCP is installed and ready to use.

Adding Your API Key

Browser-Use MCP requires an API key for access. This key connects your computer to the Browser-Use Cloud service.

Go to Browser Use Cloud and create your API key. New users get $10 in free credits to start.

After you get your key, add it to a file called .env in your project folder like this:

# .env
BROWSER_USE_API_KEY=your-key

Replace your-key with the actual key you copied. This step is important for the tool to work.

Installing Chromium Browser

Browser-Use MCP needs a browser to control. You can install the Chromium browser by running this command:

uvx browser-use install

This sets up the actual browser software that Browser-Use MCP will automate.

Running Your First Agent

Now that you have everything installed, let’s run your first simple script with Browser-Use MCP.

Here’s the example code you can use:

from browser_use import Agent, Browser, ChatBrowserUse
import asyncio

async def example():
    browser = Browser(
        # use_cloud=True,  # Uncomment this line to use a stealth browser on Browser Use Cloud
    )

    llm = ChatBrowserUse()

    agent = Agent(
        task="Find the number of stars of the browser-use repo",
        llm=llm,
        browser=browser,
    )

    history = await agent.run()
    return history

if __name__ == "__main__":
    history = asyncio.run(example())
    print(history)

This script creates a browser and an AI agent that will look up how many stars the Browser-Use GitHub repo has. You can run it in your Python environment.

Extra Tips and Examples

Browser-Use MCP also offers ready-to-run templates to get started faster. You can generate one by typing:

uvx browser-use init --template default

This creates a file with a working example. Other templates like advanced or tools are also available for more features.

If you want to explore more sample projects, check the official examples page here.

---

By following these steps, you’ll quickly install Browser-Use MCP and start automating browser tasks in no time! It’s a powerful yet easy-to-use tool for both beginners and developers.