Apify MCP: Automate Web Scraping with Claude [2026]

Stop copy-pasting text from websites. Learn how to install and use the Apify MCP to extract real-time data directly into your AI workflows securely and automatically.

Quick Answer

The Apify MCP connects Claude Desktop to Apify's cloud platform of over 2,000 scraping "Actors". By configuring the `@apify/mcp-server` with your `APIFY_TOKEN`, you can type "Claude, scrape the pricing page of competitor X" and the system will bypass bot protections, extract the data, and summarize it in your chat.

⏱️ TL;DR: Cloud-based web scraping directly inside Claude.

Why Use Apify Instead of a Local Fetch tool?

Standard MCP fetch tools (like `@modelcontextprotocol/server-fetch`) simply perform a basic `cURL` request from your local IP. In 2026, 90% of modern websites block basic HTTP requests with Cloudflare, Datadome, or CAPTCHAs. Furthermore, local fetch tools cannot render React/Vue Single Page Applications.

Apify solves this. It executes the request on a headless cloud browser utilizing proxy rotation, processes the JavaScript, extracts the raw data, and passes the clean JSON back to Claude.

Tutorial: Setting Up Apify MCP

Prerequisites

An Apify account (the free tier works fine).
Your personal `APIFY_TOKEN`.
Claude Desktop App installed.

Step 1: Configure Claude Desktop

Add the Apify server to your `claude_desktop_config.json` via npx.

{
  "mcpServers": {
    "apify": {
      "command": "npx",
      "args": ["-y", "@apify/mcp-server"],
      "env": {
        "APIFY_TOKEN": "apify_api_VyourTokenHere123"
      }
    }
  }
}

Step 2: Tell Claude to Run an Actor

Restart Claude. You now have access to the `run_actor` tool. Simply tell Claude what you want:

"Claude, please use Apify to run the 'Website Content Crawler' on https://stripe.com/pricing and extract all the tiered pricing data into a markdown table."

Claude will automatically find the correct Actor ID (`apify/website-content-crawler`), format the JSON input payload, trigger the cloud run, wait for completion, and stream the results back to you.

✅ Do:

• Tell Claude exactly which Actor name to search for (e.g., "Google Maps Scraper").
• Ensure your `APIFY_TOKEN` is kept strictly within the env object.

❌ Don't:

• Don't use Apify for scraping 10,000 pages at once via Claude chat (it will hit Claude's token limits). Scrape specific targeting pages.

Frequently Asked Questions

Is the Apify MCP free?

The MCP server software itself is open-source and free. However, running Apify Actors (the actual cloud scraping bots) consumes Apify platform credits. Apify offers a generous free tier that covers basic personal scraping tasks.

Can Claude bypass CAPTCHAs with Apify?

Yes, indirectly. Because Claude triggers an Apify Actor in the cloud, that Actor uses Apify's residential proxies and sophisticated browser fingerprinting to bypass anti-scraping measures like Cloudflare or Datadome, returning the clean JSON to Claude.

Connect APIs Automatically

Apify is just one of many API integrations. Browse our directory to find MCPs for Google Drive, Jira, Notion, and more.

Explore Integrations