Puppeteer MCP server

Tired of fragile scraping scripts? This MCP server connects AI agents directly to Puppeteer's headless Chrome automation for reliable web scraping, form filling, and data extraction—no maintenance headaches.

Try Puppeteer MCP server

POWERING THE WORLD'S TOP DATA-DRIVEN TEAMS

What is Puppeteer MCP server?

Puppeteer MCP server connects AI agents with Puppeteer scraper via the Model Context Protocol. It fetches web pages with JavaScript execution, extracts data through custom page functions, handles link crawling with CSS selectors, manages proxy configurations, and more.

Full browser automation with JavaScript

Uses Puppeteer library to control a headless Chrome browser programmatically, enabling complex interactions and dynamic content scraping that simpler tools can't handle.

Recursive website crawling capabilities

Automatically discovers and follows links using CSS selectors, glob patterns, and pseudo-URLs to crawl entire websites systematically without manual URL specification.

Advanced data extraction with Node.js power

Provides complete access to Puppeteer's Page API and Node.js environment through customizable page functions, allowing sophisticated data processing and extraction logic.

One token to rule them all

Single Apify API key unlocks 6,000+ Actors and MCP servers. One account, one invoice and transparent usage tracking.

Dynamic tool discovery

Is Puppeteer MCP server not enough? Agents can search, evaluate, and add other Apify tools based on workflow needs.

Smart prompts and resources

Pre-built prompts and resources help agents understand tool capabilities and use cases.

Prerequisites

Before you start, make sure you have the following:

An Apify account

Apify API Token

Get your personal API token from the Integrations section in Apify Console. This token will be used to authorize the MCP server to run Actors on your behalf.

MCP client

An AI agent or client that supports MCP. This could be Anthropic Claude for Desktop, a VS Code extension with MCP support, Apify’s web-based Tester MCP Client, or any custom client implementation. You can also check supported MCP clients in official documentation.

How to set up Puppeteer MCP server

If you have an MCP client that supports streamable HTTP, connect using OAuth authentication by providing Puppeteer MCP server URL you can find below. You'll be redirected to your browser to sign in to your Apify account and approve the connection.

1{
2 "mcpServers": {
3   "apify": {
4     "url": "https://mcp.apify.com/?actors=apify/puppeteer-scraper"
5   }
6 }
7}

Check documentation

If you don't have an MCP client that supports streamable HTTP, you can still set up Apify's Puppeteer MCP server. For Claude for Desktop, enable MCP servers via the Developer settings. Then, you need to add an entry for the Apify MCP server. For instance, in Claude’s config file, under mcpServers, add an entry like:

{
    "mcpServers": {
      "actors-mcp-server": {
        "command": "npx",
        "args": [
          "-y", "@apify/actors-mcp-server",
          "--actors", "apify/puppeteer-scraper"
        ],
        "env": {
           "APIFY_TOKEN": "YOUR_APIFY_TOKEN"
        }
      }
    }
   }

Check documentation

MCP server configuration

Configure your own Apify MCP server exactly as you need with Actors and tools for easy integration with MCP clients.

Task examples for Puppeteer MCP server

"Take a screenshot of the homepage of reddit.com"

"Navigate to amazon.com and search for wireless headphones"

"Click on the login button and fill out the username field with testuser"

"Scroll down to the bottom of the page and extract all footer links"

Sample dataset

Here's a sample of the data you can extract with our Puppeteer MCP server including web scraping data, form field information, automated interaction results, and extracted content from headless Chrome sessions, and much more without the limitation of traditional web platform's official API.

{
  "title": "Web Scraping, Data Extraction and Automation - Apify",
  "#error": false,
  "#debug": {
    "requestId": "fvwscO2UJLdr10B",
    "url": "https://apify.com",
    "loadedUrl": "https://apify.com/",
    "method": "GET",
    "retryCount": 0,
    "errorMessages": null,
    "statusCode": 200
  }
}

Try Puppeteer MCP server online

Use the Tester MCP Client, our chat user interface, to interact with Puppeteer MCP server online. And for free.

Tester MCP Client

jiri.spilka/tester-mcp-client

A model context protocol (MCP) client that connects to any MCP server using Streamable HTTP and displays the conversation in a chat-like UI. It is a standalone Actor server designed for testing MCP servers over Stremable HTTP.

Jiří Spilka

847

5.0

Frequently asked questions

The Puppeteer MCP server is a tool that enables AI agents and applications to extract text content, form data, page screenshots, and navigation elements from websites using headless Chrome automation for reliable web scraping and data extraction. It provides programmatic access to comprehensive data beyond the limitations of traditional web scraping's official API constraints.

To use the Puppeteer MCP server, create an Apify account and get your API token. Then configure the server in your MCP-compatible client using either OAuth authentication or by adding the configuration with your token.

No, you don't need Puppeteer API key to use our Puppeteer MCP server. The server operates through Apify's infrastructure and only requires an Apify account with API token for authentication and access to the service.

You can access a wide range of web data through Puppeteer MCP server's headless Chrome automation, including text content, form data, images, links, and structured elements from any webpage. The server enables reliable extraction of dynamic content, user interactions, and automated form submissions without the typical maintenance issues of traditional scraping scripts.

The Puppeteer MCP server has no specific rate limits imposed by the service itself, but usage is subject to your Apify account's resource limits and pricing plan.