Puppeteer MCP
Browser automation and web scraping for AI
How to Install Puppeteer MCP
$npx @modelcontextprotocol/server-puppeteerRequires Claude Desktop, Cursor, Windsurf, or another MCP-compatible client.
About Puppeteer MCP
Drive a headless Chrome from your AI agent. Click, type, screenshot, extract.
Puppeteer MCP is a Browser Automation MCP server designed for QA, growth, ops, and research teams. It enables driving a real browser from an AI assistant by exposing Playwright/Puppeteer sessions as MCP tools. Key capabilities include navigate, click, type, and submit, headless screenshots and pdfs, form-fill and file-upload flows, and structured tool calls that work cleanly inside an LLM context. It integrates with Playwright, Puppeteer, headless Chromium, and Selenium grids, and is best suited for QA, growth, ops, and research teams who need end-to-end testing.
Key Features
- browser
- automation
- scraping
- Navigate, click, type, and submit
- Headless screenshots and PDFs
- Form-fill and file-upload flows
Pricing
- Core MCP server
- Community support
- Works with any MCP client
- Everything in Free
- Higher usage limits
- Priority support
- Everything in Pro
- SLA & SSO
- Dedicated support
Tier details are indicative — visit the Puppeteer MCP website for current pricing.
Pros & Cons
Pros
- Reaches sites that don't expose APIs
- Self-healing test flows
- Strong selectors paired with LLM context
Cons
- Slower than direct API calls
- Brittle against aggressive bot protection
- Per-session state can leak without isolation
Best For
- End-to-end testing
- Scraping rendered pages
- Form-heavy workflows
- Dashboard data extraction
Screenshots
Screenshots coming soon —
Submit yours →Our Take on Puppeteer MCP
Best-in-class browser control for AI. Watch out for memory usage on long sessions.
Alternatives to Puppeteer MCP
Fetch any URL or web content from AI
See how Puppeteer MCP compares to Fetch MCP →