Skip to main content

Crawl with Browser Use and Blitzbrowser

In this tutorial, we will see how we can use Browser Use and BlitzBrowser as a AI browser agent.

This stack is perfect when you want to automate repetitive work and not manage multiple browsers. Use the power of AI to automate your browser tasks.

Requirements for this tutorial​

Prepare your environment​

Create a new directory named blitzbrowser-browser-user where you will run this tutorial.

Browser Use is easy to install on your machine. You have to follow their quickstart documentation and you will be ready in few minutes.

If you don't have an account, you need to create an OpenAI account to run the code. Once you have an OpenAI API key, you have to set the environment variable OPENAI_SECRET_KEY. If you want to use another LLM, you can follow Browser Use documentation here.

Create the main.py file and run the insert the following code.

from langchain_openai import ChatOpenAI
from browser_use import BrowserSession, Agent
import os
import asyncio

# Browser config to use BlitzBrowser browsers
browser_session = BrowserSession(cdp_url=f"ws://localhost:9999")

async def main():
    agent = Agent(
        task="Go to https://docs.blitzbrowser.com, find their documentation and create a summary of each doc into a JSON array. Once the JSON list extracted you can stop your task.",
        # LLM config to connect to OpenAI
        llm=ChatOpenAI(
            model="gpt-4o",
            api_key=os.environ.get('OPENAI_SECRET_KEY')
        ),
        browser_session=browser_session
    )
    result = await agent.run()
    print(result.extracted_content())

asyncio.run(main())

Run your AI browser agent​

This example contains everything to run Browser Use and BlitzBrowser out-of-the-box. This example will go on docs.blitzBrowser.com, find all the documentation and return the results in a JSON array.

Conclusion​

You are now be ready to automate any web task with Browser Use and BlitzBrowser.