Practical overview
When it fits
Use it when an MCP-compatible AI client needs access to ready-made Apify scrapers, crawlers or automation Actors instead of implementing each website-specific extraction workflow directly.
Who it is for
- Developers building MCP-compatible AI assistants.
- Teams creating web research or structured data-extraction workflows.
- Builders of market-monitoring, social-media analysis or location-intelligence agents.
Documented uses
- Extract data from social media, search engines, maps, e-commerce sites and other websites.
- Invoke ready-made Apify Actors as tools from an AI assistant.
- Fetch website content as Markdown, plain text, HTML or links.
- Connect a hosted MCP endpoint or configure a local stdio integration.
Requirements & limits
The documented hosted endpoint is https://mcp.apify.com and supports OAuth or an Apify token. A local stdio option uses npx @apify/actors-mcp-server and the APIFY_TOKEN environment variable. The repository is public and MIT licensed.
- Hosted access supports output-schema inference for structured Actor results; the documented local stdio route does not.
- The client must adhere to MCP, and support for dynamic tool discovery may vary.
- The current documented hosted transport is streamable HTTPS; the legacy /sse endpoint has been removed.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Connect an MCP client to the hosted Apify endpoint
Expose Apify Store scrapers, crawlers and automation tools to an MCP-compatible AI client.
Before you start
- An MCP-compatible client such as Claude.ai or Visual Studio Code.
- The hosted endpoint: https://mcp.apify.com.
- OAuth or an Apify API token for authorization.
Steps
Open the MCP client configuration or connection flow.
Add https://mcp.apify.com as the MCP server URL.
Authorize with OAuth, or configure requests with an Authorization: Bearer header.
Use the connected assistant to invoke available Apify Actors for the required extraction task.
Expected outcome
The AI client can invoke Apify website data-extraction and automation capabilities through MCP.
Guide limitations
- Support for dynamic tool discovery and other MCP features may vary between clients.
- The documented endpoint uses streamable HTTPS; the legacy /sse endpoint has been removed.
Configure a local standard-input/output integration
Set up the documented local MCP server command for a command-line client.
Before you start
- A local MCP-compatible client, such as Claude for Desktop.
- The @apify/actors-mcp-server package available through npx.
- An Apify API token stored in the APIFY_TOKEN environment variable.
Steps
Set the MCP client server command to npx @apify/actors-mcp-server.
Provide the APIFY_TOKEN environment variable with the Apify API token.
Use npx @apify/actors-mcp-server --help for the documented command options.
Connect the client to the local MCP server.
Expected outcome
A local MCP client is configured to use the Apify MCP server through standard input/output.
Guide limitations
- The documentation states that output-schema inference for structured Actor results is available on the hosted server but not through local stdio.
- The supplied documentation does not provide a deployment test result.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 1 October 2026
- Original website title
- apify/apify-mcp-server · Official homepage title · retained in its source language