google-maps-mcp-server

Google Maps MCP Server

A standalone Model Context Protocol server that exposes Google Maps business search as a single MCP tool, backed by Techforce Global's Google Maps Scraper Apify Actor.

Tool

search_google_maps

Search Google Maps for businesses/places by keyword and location.

Parameter	Type	Required	Description
searchQueries	string[]	yes	One or more search terms, e.g. "restaurants in Delhi"
maxResults	integer	no	Max places per query (default 20, max 500)
language	string	no	Language code for results, e.g. "en"

Returns structured JSON: business name, address, phone, rating, opening hours, and GPS coordinates.

Authentication

Every request must include your own Apify API token:

Authorization: Bearer <YOUR_APIFY_TOKEN>

Usage is billed to the token owner's Apify account, not to Techforce Global. Get your token at Apify Console → Settings → Integrations.

Installation

Hosted (recommended)

Connect directly to the hosted endpoint — no install required:

https://<your-vercel-project>.vercel.app/mcp

Example client config (Claude Desktop / any MCP client supporting Streamable HTTP):

{
  "mcpServers": {
    "google-maps": {
      "url": "https://<your-vercel-project>.vercel.app/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_APIFY_TOKEN>"
      }
    }
  }
}

Run locally

npm install
npm start

Server listens on http://localhost:4000/mcp.

Deploying on Vercel (free)

This project includes a serverless entry point (api/mcp.js) and a vercel.json rewrite so the public endpoint is available at a clean /mcp path.

1. Go to vercel.com, sign in, click Add New → Project.
2. Import this GitHub repo (Sagar27888/google-maps-mcp-server).
3. Framework preset: Other. No build command needed — Vercel auto-detects the api/ folder.
4. Click Deploy.
5. The public endpoint becomes available at https://<your-vercel-project>.vercel.app/mcp.

Deploying on Apify (alternative)

This project is also an Apify Actor configured for Standby mode, which turns it into a long-running HTTP service instead of a one-shot batch job.

1. Push this project to an Actor on your Apify account (Apify Console → Actors → Create new → connect this source, or use the Apify CLI: apify push).
2. In the Actor's Settings → Standby, enable Standby mode.
3. The public endpoint becomes available at https://<actor-name>.<your-apify-username>.apify.actor/mcp.

License

MIT