google-maps-transit
A Model Context Protocol (MCP) server for Google Maps — routing, place discovery, and commute comparison over stdio.
README
maps-mcp
A Model Context Protocol server for Google Maps — routing, place discovery, and commute comparison over stdio.
Tools
get_transit_commute
Route between two addresses (or neighborhoods). Supports transit (default), driving, walking, bicycling, two_wheeler. Multi-leg trips via intermediates (A→B→C). Returns up to 3 alternative routes, each with:
mode,duration_minutes,distance_km,summary,steps_count- transit:
segments(line + vehicle type),transfers,walking_minutes,first_departure,last_arrival - driving / two_wheeler:
duration_in_traffic_minutes(traffic-aware),toll_roads
search_places
Find restaurants, hotels, attractions, parks, gas stations, etc. using free-text queries. Optionally bias by location (near), filter by place type (type), and filter to currently-open places (open_now). Returns up to 20 results with name, address, rating, price level, open/closed status, and a id for follow-up.
get_place_details
Full details for a place from search_places: phone, website, per-day opening hours, editorial summary, and up to 5 user reviews. Takes a place_id.
compare_commutes
Rank 2–5 candidate origins against a shared destination by travel time. Fans out in parallel and returns results sorted fastest → slowest. Per-origin failures surface inline rather than aborting the whole comparison.
Google Cloud setup
In Google Cloud Console, pick or create a project and enable these APIs (APIs & Services → Library):
- Routes API — for
get_transit_commuteandcompare_commutes - Places API (New) — for
search_placesandget_place_details
Then Credentials → Create credentials → API key. Restrict the key to those two APIs. Put the value in .env:
echo 'GOOGLE_MAPS_API_KEY=your-key-here' > .env
Setup & running
npm install
npm run build
npm start # or: npx tsx src/index.ts (dev, no build needed)
| Make target | What it does |
|---|---|
make install |
npm ci |
make build |
Compile TypeScript → dist/ |
make dev / make start |
Run server (dev / compiled) |
make test |
Unit + in-memory e2e tests |
make test-live |
Live API test (needs key) |
make smoke |
One real API request |
make docker-build / make docker-run |
Build/run Docker image |
Connecting a client (e.g. Cursor / Claude Desktop)
{
"mcpServers": {
"maps-mcp": {
"command": "node",
"args": ["/absolute/path/to/maps-mcp/dist/index.js"],
"env": { "GOOGLE_MAPS_API_KEY": "your-key-here" }
}
}
}
For dev without building: "command": "npx", "args": ["tsx", "/absolute/path/to/maps-mcp/src/index.ts"].
Notes
- MCP server name:
google-maps-transit drivingandtwo_wheeleruseTRAFFIC_AWARErouting (Routes API v2)- Up to
MAX_ROUTES = 3alternatives returned; Google may return fewer - Missing
GOOGLE_MAPS_API_KEYsurfaces as a clear tool error
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。