Repolode
Submit
apify

apify/apify-mcp-server

An MCP server that lets AI agents run any Apify Store scraper or automation Actor to extract data from social media, search engines, maps, and websites.

View on GitHubWebsite
Stars
15,060
Forks
436
Open issues
244
Last commit
24 hours ago

What is it?

Apify MCP Server is a Model Context Protocol server that gives AI assistants access to Apify Actors, the ready-made scrapers, crawlers, and automation tools in the Apify Store. An agent can search the Store, read an Actor's input schema, run it, and retrieve the resulting dataset, so tasks such as pulling Google Maps listings, Instagram posts, or SERP results become tool calls rather than custom scraping code.

The hosted endpoint at mcp.apify.com supports OAuth and Streamable HTTP, so clients like Claude.ai and VS Code can connect using only the URL. A stdio version runs locally via npx for clients such as Claude Desktop. Tools are grouped into categories (actors, docs, runs, storage, tasks, schedules, builds, versions, api) that can be selected individually through the tools parameter.

The server also covers management tasks on the Apify account, including managing saved tasks, schedules, builds, and Actor versions, and it offers agentic payment options (AGI tokens, direct x402 USDC payments, and Skyfire) for running Actors without a standard API token.

Key features

  • 01Lets agents search Apify Store and call any Actor as a tool using its input schema.
  • 02Hosted server at mcp.apify.com with OAuth and Streamable HTTP support.
  • 03Local stdio mode launched with npx @apify/actors-mcp-server.
  • 04Tool categories can be enabled selectively via the tools parameter.
  • 05Includes Apify docs search, dataset and key-value store access, and run logs.
  • 06Supports agentic payments through AGI tokens, direct x402 USDC, or Skyfire.

Use cases

  • Letting Claude or Cursor pull Google Maps, Instagram, or SERP data during a research task.
  • Giving an agent a fetch-and-markdown tool for arbitrary web pages with JavaScript rendering.
  • Managing Apify Actor runs, schedules, and datasets from within an AI assistant.
  • Testing and debugging MCP tool integrations with the Apify Tester client.

Quick start

  1. 1

    Connect your MCP client to the hosted server by adding the URL

  2. 2

    Run the stdio server locally with npx (set APIFY_TOKEN to your Apify API token first)

    npx @apify/actors-mcp-server

Things to know

  • Full features such as output schema inference are only available on the hosted server, not in local stdio mode.
  • Most usage requires an Apify API token or a paid agentic payment method.

In collections

Open-source AI agent frameworks & appsMCP servers

Topics

#mcp#web-scraping#ai-agents#data-extraction#automation#llm-tools#agents#ai#mcp-server

FAQ

What is apify-mcp-server?
An MCP server that lets AI agents run any Apify Store scraper or automation Actor to extract data from social media, search engines, maps, and websites. Apify MCP Server is a Model Context Protocol server that gives AI assistants access to Apify Actors, the ready-made scrapers, crawlers, and automation tools in the Apify Store. An agent can search the Store, read an Actor's input schema, run it, and retrieve the resulting dataset, so tasks such as pulling Google Maps listings, Instagram posts, or SERP results become tool calls rather than custom scraping code.
Is apify-mcp-server free to use? What license does it have?
apify-mcp-server is open source under the MIT license. Check the license terms before using it in commercial products.
Is apify-mcp-server still maintained?
Actively maintained. The last commit to apify-mcp-server was on Nov 6, 2026.
What can I use apify-mcp-server for?
Letting Claude or Cursor pull Google Maps, Instagram, or SERP data during a research task. Giving an agent a fetch-and-markdown tool for arbitrary web pages with JavaScript rendering. Managing Apify Actor runs, schedules, and datasets from within an AI assistant. Testing and debugging MCP tool integrations with the Apify Tester client.
How do I install apify-mcp-server?
Connect your MCP client to the hosted server by adding the URL Run the stdio server locally with npx (set APIFY_TOKEN to your Apify API token first): npx @apify/actors-mcp-server
Where can I get apify-mcp-server?
The source code and installation instructions are on GitHub: https://github.com/apify/apify-mcp-server

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects