An MCP server for nearby place searches with IP-based location detection.
https://github.com/kukapay/nearby-search-mcpYou know the drill: every time you need location-aware search in your applications, you're stuck dealing with geolocation permissions, coordinate parsing, or asking users where they are. The Nearby Search MCP server cuts through that friction entirely.
This MCP server automatically detects location from IP addresses and pipes that directly into Google Places searches. No permission prompts, no coordinate guessing, no user input required. Just ask for "coffee shops" and get results ranked by actual distance from the current location.
# Instead of this complexity:
user_coords = get_user_location() # Permission prompt, error handling, etc.
places = search_places(query="coffee", lat=user_coords.lat, lng=user_coords.lng)
# You get this simplicity:
places = search_nearby(keyword="coffee")
AI Assistant Integration: Your Claude Desktop setup can now provide location-aware recommendations without any user friction. Ask "find me a good sushi place" and get actual nearby results, not generic suggestions.
Automated Location Intelligence: Perfect for travel planning tools, expense apps, or any automation that needs to know what's actually around you. No more hardcoded cities or manual location entry.
Rapid Prototyping: Testing location-based features becomes instant. Drop this into your MCP setup and start building location-aware functionality without the usual geolocation setup overhead.
Business Travel Workflows: Automatically find restaurants, coffee shops, or services near your current location when traveling. Perfect for expense tracking apps that need to know what's actually available nearby.
Content Creation: Travel bloggers or location-based content creators can quickly research what's around them without switching between multiple apps or manually entering locations.
Research and Analysis: Market researchers can instantly understand the competitive landscape around any IP location - what restaurants, services, or businesses are actually nearby.
The server exposes a single search_nearby tool with three parameters:
keyword: What you're looking for ("italian restaurant", "gas station")radius: Search distance in meters (defaults to 1500m)type: Optional Google Places type filter ("restaurant", "cafe", "hospital")Setup requires just a Google Places API key. The server handles IP geolocation automatically using ipapi.co, so there's no additional location service configuration needed.
Add this to your MCP configuration and you'll have location-aware search capabilities in any MCP-compatible environment:
{
"mcpServers": {
"nearby-search": {
"command": "uv",
"args": ["--directory", "path/to/nearby-search-mcp", "run", "main.py"],
"env": {
"GOOGLE_API_KEY": "your_google_api_key"
}
}
}
}
The beauty is in what you don't have to build: no geolocation handling, no place search infrastructure, no coordinate management. Just clean, location-aware results when you need them.