Skip to main content
Deploy OpenBrowser with saved authentication using browser profiles.

​
1. Test Locally with Existing Chrome Profile

Use your existing Chrome profile with saved logins: 
from openbrowser import Agent, Browser, ChatGoogle

browser = Browser(
    executable_path='/Applications/Google Chrome.app/Contents/MacOS/Google Chrome',
    user_data_dir='~/Library/Application Support/Google/Chrome',
    profile_directory='Default',
)
agent = Agent(task="your task", browser=browser, llm=ChatGoogle())
await agent.run()
Close Chrome completely before running to avoid conflicts.

​
Chrome Profile Paths by OS

OSUser Data Dir
macOS~/Library/Application Support/Google/Chrome
Windows%LOCALAPPDATA%\Google\Chrome\User Data
Linux~/.config/google-chrome

​
2. Use Persistent Browser Session

For production, use a dedicated user data directory: 
from openbrowser import Agent, Browser, ChatGoogle
import tempfile

# Create persistent browser with custom user data
browser = Browser(
    user_data_dir='./browser_data',  # Persists between runs
    headless=False,  # Set True for production
)

agent = Agent(
    task="Login to example.com and check notifications",
    browser=browser,
    llm=ChatGoogle()
)
await agent.run()

​
3. Remote Browser via CDP

Connect to a remote browser using Chrome DevTools Protocol: 
from openbrowser import Agent, Browser, ChatGoogle

# Connect to existing Chrome instance
# Start Chrome with: chrome --remote-debugging-port=9222
browser = Browser(cdp_url="http://localhost:9222")

agent = Agent(
    task="your task",
    browser=browser,
    llm=ChatGoogle()
)
await agent.run()

​
Best Practices

  1. Use environment variables for API keys
  2. Run headless in production (headless=True)
  3. Set allowed_domains to restrict navigation
  4. Use proxy for rate limiting and geo-targeting
  5. Monitor with telemetry for debugging
from openbrowser import Agent, Browser, ChatGoogle

browser = Browser(
    headless=True,
    allowed_domains=['*.example.com'],
    proxy={'server': 'http://proxy:8080'},
)

agent = Agent(
    task="Production task",
    browser=browser,
    llm=ChatGoogle(),
)