Integrations / Automation libraries / browser-use
browser-use in a profile of its own.
browser-use drives a browser with an LLM. Give it a Browz profile instead of its own Chromium and each run gets its own fingerprint, proxy and cookies, in a window you can watch and take over.
pip install browser-use # plus the Browz Python SDK: pip install browz-mcpSetup
Set it up.
Install Browz and open it
Download Browz on the computer that runs the agent.
Install
Terminalpip install browser-use # plus the Browz Python SDK: pip install browz-mcpRun an agent in a fresh profile
Set
OPENAI_API_KEYfirst.browser_use_browser(session)is browser-use’sBrowser(cdp_url=…)pointed at the profile.agent.pyimport asyncio from browser_use import Agent, ChatOpenAI from browz.agents import browz_session, browser_use_browser async def main(): with browz_session() as session: agent = Agent(task="Open https://example.com and tell me the page title", llm=ChatOpenAI(model="gpt-5"), browser=browser_use_browser(session)) print((await agent.run()).final_result()) asyncio.run(main())
API from browser-use remote browser docs.
Example
Keep the profile for next time.
With keep=True the profile, and any sign-in you completed in it, stays. Next run, pass profile_id to come back to it.
import asyncio
from browser_use import Agent, ChatOpenAI
from browz.agents import browz_session, browser_use_browser
async def main():
with browz_session(name="browser-use-research", keep=True) as session:
print("profile id:", session.id) # pass profile_id=... next time
agent = Agent(task="Open https://example.com and summarise it in one sentence",
llm=ChatOpenAI(model="gpt-5"), browser=browser_use_browser(session))
print((await agent.run()).final_result())
asyncio.run(main())Good to know
Notes and tips.
- Browz starts the browser, so the profile’s own window size, proxy and fingerprint take the place of browser-use’s launch options.
- Attach to the browser Browz started; never let the library launch its own. A library-launched browser is a stock one with none of the profile’s fingerprint, proxy or cookies.
- A person completes sign-ins, 2FA codes and CAPTCHAs in the profile’s window; your script can wait for the page that follows. Browz never solves or bypasses CAPTCHAs.
- Automate only where each site’s terms allow it. Our acceptable use policy applies to scripts the same as to people.
- The local API, CDP and the SDKs are on every plan, Free included (5 profiles), and are not metered.
FAQ
browser-use and Browz.
Do I need the Browz MCP server for this?
No. browser-use talks to the browser over CDP; the SDK only starts and stops the profile through the local API.
Which models work?
Any model browser-use supports. The example uses OpenAI; swap the llm for another provider.
Can I watch the agent?
Yes. The profile opens as an ordinary window unless you start it headless, and you can take over at any time.
Get started
Download Browz. Every task, its own browser.
curl -fsSL https://gobrowz.com/agents/install.sh | sh