Provides access to Apify Actors as tools for data extraction and automation via MCP clients.
Configuration
View docs{
"mcpServers": {
"apify-apify-mcp-server": {
"url": "https://mcp.apify.com",
"headers": {
"APIFY_TOKEN": "YOUR_APIFY_TOKEN",
"APIFY_META_ORIGIN": "STANDBY"
}
}
}
}The Apify MCP Server lets your AI agents use any Apify Actor as a tool to perform tasks across websites and data sources. It enables dynamic tool discovery, easy integration with clients, and access to thousands of Actors for data extraction, automation, and web interactions.
Connect your MCP client to the hosted server at https://mcp.apify.com to leverage OAuth or provide an Authorization: Bearer <APIFY_TOKEN> header in requests. This lets you run Actors from the Apify Store as tools in your conversations. You can use clients like Claude Desktop, Visual Studio Code, or VS Code Genie, and Apify Tester MCP Client, to interact with the server and discover tools such as search, run, and data retrieval Actors. When you request an Actor-based tool, the server selects the appropriate tool set for your client’s capabilities, enabling dynamic discovery where supported.
You can start using the MCP Server in two ways: the hosted HTTPS endpoint or a local stdio setup. For most scenarios, the hosted endpoint is easiest and supports OAuth.
If you want to run a local MCP server via stdio, you can use the following CLI configuration to load the same toolset as the hosted server. This exposes a set of Actor tools and helper functions for testing and development.
npx @apify/actors-mcp-server --tools actors,docs,apify/rag-web-browserBefore running locally, ensure you have Node.js v18 or higher installed. Then install dependencies and start the server as shown above. If you plan to connect to the hosted server, you mainly need your Apify API token for authentication.
If you encounter issues, verify your Node.js version and the presence of your Apify API token. Use the MCP Inspector for debugging local stdio runs when you need deeper inspection.
The hosted server supports OAuth for convenient client integrations. When using the local stdio setup, pass the necessary credentials (such as your Apify API token) through environment variables or the tool configuration to ensure secure access.
Leverage tools like the web browser Actor for RAG-style browsing, or search and fetch tools to query Actors and retrieve outputs. You can also use the get-actor-run series of tools to monitor your Actor executions and retrieve logs for debugging.
Search for Actors in the Apify Store.
Retrieve detailed information about a specific Actor.
Call an Actor and obtain its run results.
An Actor tool to browse the web and gather content.
Search the Apify documentation for relevant pages.
Fetch the full content of an Apify documentation page by URL.
Get detailed information about a specific Actor run.
Get a list of an Actor’s runs with filtering by status.
Retrieve logs for a specific Actor run.
Get metadata about a specific dataset.
Retrieve items from a dataset with filtering and pagination.
Generate a JSON schema from dataset items.
Get metadata about a specific key-value store.
List keys within a specific key-value store.
Get the value for a specific key in a key-value store.
List all available datasets for the user.
List all available key-value stores for the user.
Add an Actor as a new tool for the user to call (experimental).
Retrieve the full output from an Actor call when the preview is limited.