Connect any Apify scraper to Claude or ChatGPT in 2 minutes (MCP)

#chatgpt #mcp #ai #webscraping

You can give Claude, ChatGPT, Cursor, or any MCP client the ability to run real web scrapers and pull live data — search results, web pages behind Cloudflare, business leads, job listings — without writing any integration code. Here's how, using the Apify MCP server.

The one thing you need to know

Apify hosts an MCP server at https://mcp.apify.com. Point any MCP client at it and your AI can discover and run any of thousands of Apify Actors (scrapers), read their results, and adapt to new tasks on the fly.

Claude Desktop (easiest)

Settings → Connectors → Add custom connector
Server URL: https://mcp.apify.com
Your browser opens once to sign in to Apify and approve access.

That's it — ask Claude "search the web for X" or "scrape this URL" and it runs the tool.

Any MCP client (Cursor, VS Code, Claude Code…)

Add this to your MCP config:

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com"
    }
  }
}

Prefer a token over OAuth? Add a header:

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com",
      "headers": { "Authorization": "Bearer <APIFY_TOKEN>" }
    }
  }
}

Get your token from the Apify Console → API & Integrations.

Load only the tools you want

By default the server loads discovery tools + a web browser. To expose specific scrapers as ready-to-call tools, list them in the URL:

https://mcp.apify.com?tools=get_anything/serp-scraper,get_anything/web-unblocker,get_anything/contact-extractor

A few of mine that are genuinely useful to LLM workflows:

serp-scraper — Google/Bing/DuckDuckGo results as JSON for RAG grounding
web-unblocker — fetch any URL past Cloudflare/DataDome and return clean Markdown
llm-web-scraper — any URL/site → LLM-ready Markdown
contact-extractor — emails/phones/socials from a list of domains
google-maps-leads — local businesses with phone, website & email

No-token discovery mode

For discovery/testing you can connect with no token by loading only the read-only tools:

https://mcp.apify.com?tools=search-actors,fetch-actor-details,search-apify-docs,fetch-apify-docs

Running an Actor (and getting data back) needs a token or OAuth sign-in.

Why this matters

Instead of copy-pasting data into your prompt, your AI fetches it live — grounded, current, and structured. That's the difference between an LLM that guesses and one that knows.

Try it: point Claude Desktop at https://mcp.apify.com and ask it to scrape something. Full docs: Apify MCP server.

DEV Community