Tools for agents · Apify Actors · MCP tools

Public data an agent can call and pay for by the row.

Each tool runs on Apify, which meters usage and bills your Apify account, and is listed on Apify's MCP server so an agent can find it by name. One row per thing, named fields, a documented source.

Catalog How to call a tool

Catalog · updated 2026-10-18 · catalog.json

First batch: public records and open data.

Statuses move from planned to live as each tool is published and approved. Prices per run and per result are set at publish and shown here and on Apify.

How to call a tool

Three ways in, same rows out.

MCP client config · exampleJSON
{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com/?tools=lathewright/company-filings",
      "headers": { "Authorization": "Bearer <your Apify token>" }
    }
  }
}
HTTP · run and return rows · examplecurl
curl -X POST "https://api.apify.com/v2/acts/lathewright~company-filings/run-sync-get-dataset-items?token=<your Apify token>" \
  -H "Content-Type: application/json" \
  -d '{ "query": "Apple Inc", "forms": ["10-K", "8-K"], "from": "2026-01-01" }'

Exact input schemas and prices live on each tool's Apify page and are linked from the catalog as tools go live. The store page is our Apify store page.

Billing

Per result, by Apify. No subscription to us.

Every tool lists a price per run start and a price per result row. Apify counts those events and bills your Apify account, the same way it bills every Actor in its store. No minimum, and no charge for rows a run did not produce.

Rules every tool follows

Request a source

Need a source we don't cover yet?

Describe it on the intake form: the source, the fields you need and how often you will call it. You get a written yes or no. If it fits the rules above, it goes on the build list and you hear when it is live.