Skip to main content
OpenClaw is a self-hosted gateway that connects chat apps like WhatsApp, Telegram, and Discord to AI coding agents. Add Browser Use and those agents get full browser automation — anti-detect profiles, CAPTCHA solving, residential proxies in 195+ countries, and stealth browsing out of the box. Two ways to set it up: connect a Browser Use cloud browser to OpenClaw’s native browser tool via CDP, or install the Browser Use CLI as a skill.

​
Why OpenClaw gets blocked, and when you need this

  • On your own laptop, you usually don’t. OpenClaw’s local browser runs on your IP, and its user and chrome profiles drive your signed-in Chrome (OpenClaw browser profiles).
  • On a VPS or in Docker, sites see a datacenter IP and a fresh, automated Chromium with no history. Those are the usual reasons for blocks and CAPTCHAs.
  • A Browser Use cloud browser runs a hardened Chromium with stealth on, behind a residential proxy in the country you pick (US by default), with automatic CAPTCHA solving (CAPTCHA handling). This page covers the US deployment; the EU deployment has its own keys and endpoint, and does not solve CAPTCHAs automatically. On Browser Use’s Stealth Bench V2 (100 bot-protected sites, 2 visits each, September 25, 2026) it got through on 90.5% of visits; the next provider, Anchor, got 82%. The sites were picked where Browser Use Cloud historically got through and stock Chrome did not, so they are not a balanced sample, and no browser gets through everywhere.
  • Cost: $0.02 per browser-hour plus $5/GB of residential proxy traffic, no subscription (pricing).
Hosted options compared, with prices: browser-use.com/openclaw.

​
Option 1: Cloud Browser via CDP

OpenClaw has a built-in browser tool with its own CLI commands (openclaw browser). By default, it controls a local Chromium instance. You can point it at a Browser Use cloud browser instead by configuring a remote CDP profile. Browser Use exposes a WebSocket CDP URL. OpenClaw connects to it like any remote browser — no SDK or extra dependencies needed.

​
Setup

1. Get your API key Sign up at cloud.browser-use.com and copy your API key from Settings → API Keys. 2. Add a Browser Use profile Open ~/.openclaw/openclaw.json and add a browser-use profile:
Replace <BROWSER_USE_API_KEY> with your actual key. All Browser Use session parameters can be passed as query params in the cdpUrl:
  • timeout — session duration in minutes (max 240)
  • profileId — load a saved browser profile with persistent cookies and localStorage
  • proxyCountryCode — route traffic through a specific country (e.g. us, de, jp)
3. Use it OpenClaw’s browser commands now run against a Browser Use cloud browser:
If you set defaultProfile to "browser-use" in the config (as shown above), you can drop the --browser-profile flag:
If a site still blocks it
  • Try another country in proxyCountryCode. Some sites block whole regions or IP ranges.
  • For sites you log in to, save the login in a profile once and add &profileId=<PROFILE_ID> to the cdpUrl (profile sync). The browser starts with those cookies, so no password goes into the prompt. Sites can still expire the login, bind it to a device or network, or ask you to verify again; re-sync when that happens.
  • Slow the agent down on rate-limited sites, and use the site’s official API where one exists.

​
Option 2: Browser Use CLI

The Browser Use CLI is a standalone tool that gives any OpenClaw agent browser automation through a SKILL.md file. The agent reads the skill and learns to use the CLI commands directly. It’s available on skills.sh and ClawHub.

​
Setup

1. Install the CLI
2. Verify the installation
3. Set up the agent Paste this setup prompt into your OpenClaw agent:
Once the skill is loaded, OpenClaw agents can use the browser-use CLI to drive pages through Browser Harness and Python helpers. For the complete CLI reference, see the Browser Use CLI docs.