How to use proxies with browser-use AI agents

• An active CleanProxies plan: static ISP proxies for fixed US IPs, or rotating residential / mobile proxies for pools. The trial works too.
• Your credential line from the dashboard's proxy manager: host (or IP), port, username and password.
• Python 3.11 or later and pip install browser-use (or uv add browser-use), with Chromium installed as its README describes.
• An API key for the LLM you use with browser-use, set as an environment variable.

New to proxies? For the overview, read Proxies with Python Requests, Playwright and Puppeteer and Rotating vs sticky sessions. This guide goes deeper on browser-use only.

Most tools want the same details as a URL: http://USERNAME:PASSWORD@HOST:PORT for HTTP(S) or socks5://USERNAME:PASSWORD@HOST:PORT for SOCKS5. The examples on this page use those placeholders; copy the host, port, username and password from your CleanProxies dashboard. To convert a whole list at once, paste it into the proxy list formatter.

1. 1

   Import ProxySettings

   It isn't exported from the top-level package: use from browser_use.browser import ProxySettings.
2. 2

   Create the browser with a proxy

   Set server to http://HOST:PORT, plus username and password. bypass takes a comma-separated string such as "localhost,127.0.0.1".
3. 3

   Hand the browser to the agent

   Pass browser=browser to Agent(...) along with the task and LLM.
4. 4

   Check the exit IP

   Give the agent a first task of opening https://api.ipify.org and reporting the IP.

import asyncio
from browser_use import Agent, Browser, ChatOpenAI
from browser_use.browser import ProxySettings

browser = Browser(
    proxy=ProxySettings(
        server="http://HOST:PORT",
        username="USERNAME",
        password="PASSWORD",
        bypass="localhost,127.0.0.1",
    )
)

async def main():
    agent = Agent(
        task="Open https://api.ipify.org and report the IP address shown",
        llm=ChatOpenAI(model="gpt-4.1-mini"),
        browser=browser,
    )
    history = await agent.run()
    print(history.final_result())

asyncio.run(main())

Swap in any LLM class browser-use supports; the proxy setup doesn't change. Credentials go in the username and password fields, not in the server URL.

Give each agent its own Browser with its own proxy. With CleanProxies ISP proxies, that means one dedicated US IP per agent, so parallel agents never share an address.

def browser_for(line: str) -> Browser:
    host, port, user, password = line.strip().split(":", 3)
    return Browser(proxy=ProxySettings(server=f"http://{host}:{port}", username=user, password=password))

async def run_all(tasks, lines):
    agents = [
        Agent(task=task, llm=ChatOpenAI(model="gpt-4.1-mini"), browser=browser_for(line))
        for task, line in zip(tasks, lines)
    ]
    return await asyncio.gather(*(agent.run() for agent in agents))

• browser-use passes the server to Chromium, and Chromium supports no SOCKS5 authentication. Use the HTTP form of your CleanProxies line, or SOCKS5 without credentials on an ISP proxy authorized by IP whitelisting.
• The LLM API calls don't go through this proxy; only the browser does. That keeps your residential GB for the pages the agent visits.
• Agents retry and wander. Set a step limit and give clear tasks so a confused agent doesn't burn bandwidth on a rotating plan.

The example's task makes the agent report its own exit IP. Compare it with your dashboard line before giving the agent real work.

Open What is my IP? through the proxy to see the address and headers a website receives, or paste up to ten lines into the free proxy checker to test connectivity, latency and exit IP before you start.

Still stuck? Our 24/7 support team answers fast: contact us with the error and the line you used (leave the password out).