Set it up
-
Turn on Browserbase in your AI
In Claude: open Settings → Connectors and add Browserbase. In ChatGPT: Settings → Apps & Connectors.
-
Add it to your app once
Copy the short config below into your app. It just tells your AI where to find Browserbase.
-
Just ask
Tell it what you want in plain words — no special commands.
For example, say
“Visit https://example.com and tell me what the page title and main heading are”
Need the exact config? Show it +
Add to your client
{
"mcpServers": {
"browserbase": {
"command": "npx",
"args": ["-y", "@browserbasehq/mcp"],
"env": {
"BROWSERBASE_API_KEY": "<your-api-key>",
"BROWSERBASE_PROJECT_ID": "<your-project-id>"
}
}
}
}Paste into your client's MCP configuration file.
Get the skill
A ready-made skill that teaches your AI agent to use Browserbase well. Browserbase's official browser-automation skill.
npx skills add browserbase/skills Ready-to-run prompts
If you ask, Browserbase does it. Copy one, paste it into Claude, Codex or ChatGPT, done. No setup.
Browserbase Scrape JavaScript-heavy pages
If you ask
“Visit the SaaS pricing page at example.com/pricing and extract the plan names and prices”
Then
Read pages that require JS to load their content.
Browserbase Fill and submit forms
If you ask
“Go to the contact form on acme.com and fill it out with my name, email, and the message below”
Then
Fill out and submit forms with provided data.
Browserbase Automate a multi-step flow
If you ask
“Log into this dashboard, navigate to the reports page, and download the CSV for last month”
Then
Navigate through a flow that requires clicks and waits.
Good to know
- · Browserbase is usage-based — sessions and bandwidth add up on longer flows
- · Anti-bot measures can block cloud browsers — results vary by target site
- · Automating sites you don't own may violate their terms of service — check before running production scrapes
When to use
- ✓ The page you want to scrape requires JavaScript execution
- ✓ You need to interact with a page (click, fill, submit)
- ✓ You want browsing to happen in the cloud, not on your machine
- ✓ You're building agents that navigate multi-page flows
When not to use
- × You only need simple HTML fetches — use fetch-mcp instead (it's free)
- × You need massive parallel scraping at scale — use a dedicated scraping platform
- × You're running on a tight budget — Browserbase is usage-based
About
Imagine telling your AI: 'Go to this competitor's pricing page and tell me what they charge' or 'Fill out this web form with my data' — and it drives a real browser in the cloud to do it. This server gives your AI control of a headless Chrome browser running on Browserbase, so it can load JavaScript-heavy pages that plain HTTP fetches can't handle.
The Browserbase MCP server provides AI assistants with access to cloud-hosted browsers via the Browserbase platform. Navigate websites, interact with elements, handle authentication, bypass bot detection, and record sessions — without local browser setup.
Ideal for production web scraping, automation pipelines, and scenarios requiring consistent, scalable browser environments.
Workflows That Use Browserbase MCP
Recommended recipe
Let Claude search and browse the live webReal-time search, page scraping, and full browser automation — Claude works with what's on the web right now.
Works With
Also Consider
FAQ
What does Browserbase MCP do? +
Imagine telling your AI: 'Go to this competitor's pricing page and tell me what they charge' or 'Fill out this web form with my data' — and it drives a real browser in the cloud to do it. This server gives your AI control of a headless Chrome browser running on Browserbase, so it can load JavaScript-heavy pages that plain HTTP fetches can't handle.
Do I need to know how to code? +
No. Turn it on in your AI's settings and ask in plain English — no terminal, no coding.
When should I use Browserbase MCP? +
Reach for it when you need to: The page you want to scrape requires JavaScript execution; You need to interact with a page (click, fill, submit); You want browsing to happen in the cloud, not on your machine; You're building agents that navigate multi-page flows.
When should I avoid Browserbase MCP? +
Skip it when: You only need simple HTML fetches — use fetch-mcp instead (it's free); You need massive parallel scraping at scale — use a dedicated scraping platform; You're running on a tight budget — Browserbase is usage-based.
How do I install Browserbase MCP? +
In Claude Code, run: claude mcp add browserbase -e BROWSERBASE_API_KEY=<key> -e BROWSERBASE_PROJECT_ID=<project> -- npx -y @browserbasehq/mcp. Copy-paste configs for Claude Desktop, Cursor, and Windsurf are at the top of this page.
