DOTmed MCP Server

An MCP (Model Context Protocol) tool server that gives Claude, ChatGPT, Cursor, and any other MCP-compatible AI assistant live access to DOTmed.com's medical equipment marketplace — search, listing lookup, and price comparison — as callable tools instead of a dataset. Built on the same extraction logic as the dotmed-equipment Actor, exposed as always-ready tools via Apify Actor Standby rather than a run-once crawl.

Who uses this and for what

A procurement analyst working in Claude or ChatGPT who wants to ask "what's a fair asking price for a used Siemens CT scanner right now" without leaving the chat — the assistant calls compare_prices and answers with the actual current range, not a training-data guess.

An agentic workflow inside Cursor or a custom agent that needs to check equipment availability or price a component as one step in a larger task (a purchasing bot, a refurbishment-margin calculator) — search_equipment and get_listing give it structured tool calls instead of a scraping step it would have to build itself.

A biomedical engineer's AI assistant sourcing a specific replacement part — asks for a manufacturer-filtered search and gets back real listings with condition, price, and seller, in one tool call.

Tools

search_equipment

Search a category, optionally filtered by manufacturer. Returns up to maxResults (1–10, default 5) items with full detail.

{ "category": "dental", "manufacturer": "Planmeca", "maxResults": 5 }

get_listing

Full detail for one known DOTmed listing or auction URL (as returned by search_equipment).

{ "url": "https://www.dotmed.com/auction/dental-x-ray/planmeca/proone/282916" }

compare_prices

Samples a category (optionally filtered by manufacturer) and returns min/max/average asking price for fixed listings and the current-bid range for auctions — a quick market check before buying or selling.

{ "category": "cardiology", "sampleSize": 10 }

Real output from a live call (2026-09-26):

{
  "category": "cardiology",
  "manufacturer": null,
  "sampledItems": 5,
  "fixedListings": null,
  "auctions": { "count": 5, "minUsd": 50, "maxUsd": 11000, "avgUsd": 2270 }
}

Category slugs match dotmed-equipment's: imaging, cardiology, dental, surgical, laboratory, endoscopy, respiratory, neonatal, homecare-rehab, veterinary, ophthalmology — all verified live 2026-09-26.

Connecting an MCP client

This Actor runs in Standby mode — always ready, no run to start. Point your MCP client at its container URL with your Apify API token as a bearer token:

{
  "mcpServers": {
    "dotmed": {
      "url": "https://<your-run>.apify.actor/mcp",
      "headers": { "Authorization": "Bearer <YOUR_APIFY_API_TOKEN>" }
    }
  }
}

Pricing

Pay-per-event. Two events:

• Actor start — $0.01, charged once when a Standby instance starts (not per tool call — Standby instances serve many calls once running).
• Tool call — $0.03, charged once per completed call to any of the three tools.

Limits and honesty

• Every tool call makes one or more live requests to DOTmed.com under the house's declared identity — there is no caching between calls beyond what DOTmed itself serves fresh. A search_equipment or compare_prices call with maxResults/sampleSize near 10 makes up to 11 requests (1 category page + up to 10 detail pages) and can take several seconds.
• Same data limits as dotmed-equipment: specifications, sellerLocation, sellerType, auctionEndsAt, and listedAt are always null (see that Actor's README for why). Query-string pagination is never used (robots.txt disallows it) — a search only ever samples the first ~15 items DOTmed's own category page embeds.
• get_listing only fetches dotmed.com/listing/ or /auction/ URLs — it refuses anything else, so a client cannot use this tool as a general-purpose fetcher.
• No personal data. Same Rule 5 handling as dotmed-equipment: business names only, contact-prose patterns filtered from descriptions.

Legal

Same basis as dotmed-equipment: this server fetches only logged-out public pages that DOTmed's robots.txt permits, under the declared identity SubRosaBot/1.0, at a polite rate. See that Actor's README for the full Legal section — it applies identically here since both Actors crawl the same source the same way.

Changelog

See CHANGELOG.md.

Support

Open an issue on this Actor's Issues tab. The house replies within one day.