Google Maps MCP Server

A Model Context Protocol (MCP) server for Google Maps and local business search. Provides AI agents with tools to search for places, get business details, and find nearby locations — all using Cheerio-based scraping (no Puppeteer/Playwright required).

Features

• search_places(query, location) – Search Google Maps for businesses matching a query
• get_place_details(place_name, location) – Get detailed info including reviews, hours, and photos
• find_nearby(latitude, longitude, radius, type) – Find businesses near GPS coordinates

Quick Start

Prerequisites

• Node.js 18+
• npm

Install & Run

# Clone and install
cd mcp-google-maps
npm install

# Build TypeScript
npm run build

# Run in stdio mode (for MCP-compatible AI agents)
npm start:stdio

# Run in HTTP mode
npm start:http

MCP Integration

Claude Desktop / AI Agent Setup

Add this to your MCP client configuration:

{
  "mcpServers": {
    "google-maps": {
      "command": "node",
      "args": ["/path/to/mcp-google-maps/dist/index.js"],
      "env": {}
    }
  }
}

HTTP Mode

The server also runs as an HTTP API on port 3000:

$node dist/index.js --http

List tools:

GET http://localhost:3000/mcp/tools

Call a tool:

POST http://localhost:3000/mcp
Content-Type: application/json
{
  "tool": "search_places",
  "args": {
    "query": "coffee shop",
    "location": "New York"
  }
}

Tools Reference

search_places

Search Google Maps for places matching a query in a location.

Parameters:

Parameter	Type	Required	Description
query	string	Yes	What to search for
location	string	Yes	Where to search

Example:

{
  "tool": "search_places",
  "args": {
    "query": "pizza",
    "location": "San Francisco, CA"
  }
}

Response:

[
  {
    "name": "Tony's Pizza Napoletana",
    "address": "1570 Stockton St, San Francisco",
    "rating": 4.5,
    "ratingCount": 4723,
    "phone": "+14158359888",
    "website": "https://tonyspizza.com",
    "category": null,
    "placeId": "12345"
  }
]

get_place_details

Get detailed information about a specific place.

Parameters:

Parameter	Type	Required	Description
place_name	string	Yes	Name of the place
location	string	Yes	Where the place is located

Example:

{
  "tool": "get_place_details",
  "args": {
    "place_name": "Central Park",
    "location": "New York, NY"
  }
}

find_nearby

Find businesses near specific coordinates.

Parameters:

Parameter	Type	Required	Default	Description
latitude	number	Yes	–	Center latitude
longitude	number	Yes	–	Center longitude
radius	number	No	1000	Search radius in meters
type	string	No	"business"	Type of place to search

Example:

{
  "tool": "find_nearby",
  "args": {
    "latitude": 40.7580,
    "longitude": -73.9855,
    "radius": 500,
    "type": "restaurant"
  }
}

Apify Deployment

Deploy to Apify

# Login (if not already logged in)
apify login

# Push to Apify
apify push

The actor will be available at https://console.apify.com/actors/reverberant_equality/mcp-google-maps.

Environment Variables

Variable	Default	Description
PORT	3000	HTTP server port

Architecture

• TypeScript – Fully typed with strict mode
• Express – HTTP server for MCP transport
• MCP SDK – Model Context Protocol implementation
• Cheerio – Lightweight HTML scraping (no headless browser needed)
• Axios – HTTP client for fetching Google Maps pages
• Stdio Transport – Native MCP stdio mode for AI agent integration

Limitations

• Google Maps is heavily JavaScript-rendered; Cheerio-only scraping may not capture all results
• Results quality depends on Google's server-rendered HTML structure
• Rate limiting may apply with frequent requests
• Some fields (phone, website, hours) may not always be available

License

MIT