How To Connect OpenClaw with Chrome [Chrome MCP]

I’ll be honest with you. I hate checking emails. The endless scrolling, the spam, the newsletters I never signed up for. So I thought: why not make my AI do it?

Turns out, you absolutely can.

OpenClaw just dropped a feature that connects directly to your Chrome browser. Meaning your AI can now open tabs, read emails, and report back—all while you’re doing literally anything else.

In this post, I’ll walk you through exactly how I set it up, the costs you need to watch out for, and why this changes everything.

What Actually Happened?

I created a dummy Gmail account, sent myself a few test emails, and then asked OpenClaw via WhatsApp to read them.

Here’s the wild part: it opened my Chrome browser automatically, navigated to my inbox, read the emails, and texted me back the contents.

No clicking. No typing. Just “Hey AI, what’s in my inbox?” and boom—answers.

The Setup (It’s Embarrassingly Simple)

You ready for this? Here’s exactly what I did:

Step 1: Make sure OpenClaw is installed
Open your terminal and type:

openclaw status

If it’s not running, check out my 👉beginner tutorial to get it installed.

Step 2: Enable Chrome remote debugging
Open Chrome and go to:

chrome://inspect/#devices

Check the box that says “Enable remote debugging for this browser instance.” This lets OpenClaw talk to your browser.

Step 3: Connect your profile
In terminal, run:

openclaw browser browser-profile-user status

Chrome will pop up asking for permission. Click ALLOW.

That’s it. You’re connected.

openclaw browser --browser-profile openclaw status
openclaw browser --browser-profile openclaw start
openclaw browser --browser-profile openclaw open https://example.com
openclaw browser --browser-profile openclaw snapshot

openclaw browser --browser-profile user start
openclaw browser --browser-profile user status
openclaw browser --browser-profile user tabs
openclaw browser --browser-profile user snapshot --format ai

The Cool Part: Watching It Work

Once connected, I tested a few commands:

See all open tabs:

openclaw browser browser-profile-user tabs

Read a specific email:
I typed “read latest unread email on my inbox” in WhatsApp, hit enter, and watched OpenClaw open Gmail, find the email, and read it back to me.

Hands never touched the keyboard after that command.

The Part Nobody Talks About: Costs

Here’s where it gets interesting.

I tested this with Anthropic’s Opus 4.6 first. Reading one single email page cost me 27 cents.

Sounds small, right? But do that fifty times a day and suddenly you’re bleeding money.

So I switched to OpenAI’s GPT-4.1 mini. Same task. Same email.

Cost: Zero cents.

Literally no charge. 200,000 tokens consumed, $0 billed.

The lesson? Use the expensive models for complex stuff. Use the cheap ones for reading emails and basic browsing. Your wallet will thank me later.

Which Model Should You Use?

Here’s my honest take after testing both:

Use Opus 4.6 when:

• You’re doing complex research
• You need deep understanding
• You’re analyzing multiple sources

Use GPT-4.1 mini when:

• Reading emails
• Checking tabs
• Basic browsing tasks
• Anything you do more than ten times a day

Switching between them takes about thirty seconds. I show exactly how in the video.

Conclusion:

OpenClaw just made browser automation stupidly simple. Your AI can now do the clicking, scrolling, and reading for you. All you do is ask.

And if you’re smart about which model you use for which task, it costs next to nothing.

Want to see it in action? Watch the video above where I do the live demo, show the exact commands, and break down the costs in real time.

Got a specific model you want me to test next? Drop it in the comments. I’ll run the numbers and make another video.