Apify MCP Server

Self-host the Apify MCP Server on Railway. This gives your AI assistants (Claude, Cursor, Windsurf, etc.) access to thousands of Apify Actors for web scraping, data extraction, and automation — via the Model Context Protocol (MCP).

Deploy and Host

Deploy the Apify MCP Server on Railway with a single click. The template provisions a Node.js server that exposes Apify's MCP interface via Streamable HTTP transport, ready for any MCP-compatible client.

Dependencies for

• Apify API token — Required for authentication. Get one at console.apify.com
• MCP-compatible client — Claude Desktop, Cursor, Windsurf, or any client supporting MCP Streamable HTTP

Deployment Dependencies

Railway handles all infrastructure — no server management needed. The template uses the NIXPACKS builder with Node.js and npm dependencies declared in package.json.

About Hosting

This service runs on Railway's infrastructure using the NIXPACKS builder. It requires an Apify API token for authentication. Once deployed, Railway assigns a public HTTPS URL that serves as your MCP endpoint.

Why Deploy

• Thousands of Actors — Access the full Apify Store of scrapers and automations
• Bearer auth — Secure your endpoint with your Apify API token
• Zero configuration — Deploy with just your APIFY_TOKEN
• Always-on endpoint — Your MCP server stays available whenever you need it

Common Use Cases

• Web research — Give Claude the ability to scrape Google, Amazon, Twitter, TikTok, and hundreds more
• Data extraction — Run Apify Actors on demand via MCP tools
• AI assistant integration — Connect Claude Desktop, Cursor, Windsurf, or any MCP client
• Automated monitoring — Schedule recurring scrapes through Apify's platform

How it works

The Apify MCP Server provides a standardized MCP interface to the Apify platform. AI assistants use MCP tools to:

• Search & discover Actors in the Apify Store
• Run Actors with custom input
• Get results — retrieve and inspect Actor run outputs
• Manage datasets and scheduled tasks

This Railway deployment runs the server with Streamable HTTP transport, making it accessible to any MCP-compatible client over HTTPS.

Getting an Apify API Token

1. Sign up or log in at console.apify.com
2. Go to Settings → Integrations (or direct link)
3. Click Create token and copy the value
4. Use it as the APIFY_TOKEN variable when deploying this template

Usage

Configuring MCP Clients

Once deployed, add the server to any MCP-compatible client.

Claude Desktop

Edit ~/.claude/claude_desktop_config.json:

{
  "mcpServers": {
    "apify": {
      "url": "https://your-railway-url.up.railway.app",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Replace YOUR_APIFY_TOKEN with your actual Apify API token.

Cursor / Windsurf / Other Clients

Use the same URL and Bearer token pattern to configure your MCP client of choice.

Testing the endpoint

curl -X POST https://your-railway-url.up.railway.app/health
# {"status": "ok", "service": "apify-mcp-server"}

Environment Variables

Resources

• Apify MCP Server GitHub
• Apify Store
• Model Context Protocol