Provides access to a suite of 30 tools for OpenStreetMap data, including global geocoding, turn-by-turn routing via OSRM, point-of-interest discovery, and advanced data analysis through Overpass QL and OSMOSE quality assurance services.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@OpenStreetMap MCP Server Get driving directions from Changi Airport to Marina Bay Sands"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A powerful Model Context Protocol (MCP) server for OpenStreetMap, providing 30 tools for geocoding, routing, quality assurance, and data analysis. Optimized for AI assistants like Cursor and Claude.
A community project by Grab — providing open-source mapping tools to help developers and researchers work with OpenStreetMap data more effectively.
git clone https://github.com/GRABOSM/osm-mcp.git
cd osm-mcp
npm install
npm run buildOpen Cursor Settings → Features → Model Context Protocol.
Add a new server with:
Type: command
Command: node /absolute/path/to/osm-mcp/dist/index.js
Add this to your mcp.json:
{
"mcpServers": {
"osm-mcp": {
"command": "node",
"args": ["/absolute/path/to/osm-mcp/dist/index.js"]
}
}
}Add this to your claude_desktop_config.json:
{
"mcpServers": {
"osm": {
"command": "node",
"args": ["/absolute/path/to/osm-mcp/dist/index.js"]
}
}
}Try asking your AI:
"Find coffee shops in Makati, Manila"
"Calculate a 15-minute driving isochrone from Grab HQ Singapore"
"Get directions from Changi Airport to Marina Bay Sands"
Related MCP server: Geo MCP Worker
Geocoding: Search addresses and landmarks globally.
Routing (OSRM): Turn-by-turn directions, distance matrices, and route optimization.
Smart POI: Find restaurants, shops, and attractions with intelligent tag detection.
Quality Assurance (OSMOSE): Detect mapping errors and monitor data quality.
Changeset Analysis: Analyze recent edits and user contributions.
Tag Intelligence (Taginfo): Real-world tag usage statistics and validation.
Dual Interface: Use via MCP (AI tools) or HTTP REST API.
search_location: Search for places and addresses.
reverse_geocode: Get address from coordinates.
search_structured: Search by street, city, country, etc.
get_place_details: Get full details of an OSM element.
search_pois: General POI search (amenities, shops, etc.).
search_pois_smart: Intelligent POI search with category detection.
find_amenities_nearby: Locate facilities within a specific radius.
get_osrm_route: Detailed turn-by-turn directions.
get_distance_matrix: Travel times and distances between multiple points.
optimize_route: Find the best route through multiple stops (TSP).
map_match_gps: Snap GPS traces to the road network.
calculate_isochrone: Area reachable within a time limit.
search_highways_smart: Smart search for roads and motorways.
get_elements_in_bounds: Get all data within a bounding box.
search_by_tags: Find elements by specific OSM tags.
get_route_data: Extract road geometry for custom use.
execute_overpass_query: Run advanced Overpass QL queries.
snap_to_roads: Find the nearest road for any point.
get_changeset: Details of a specific mapping edit.
search_changesets: Find edits by user, time, or location.
get_changeset_diff: View exact changes made in an edit.
osmose_search_issues: Find data errors with filters.
osmose_get_issue_details: Details of a specific quality issue.
osmose_get_issues_by_country: Quality issues by nation.
osmose_get_issues_by_user: Issues related to a specific user.
osmose_get_stats: Data quality metrics and charts.
osmose_get_items: List of detectable issue categories.
get_tag_suggestions: Autocomplete for OSM tags.
get_tag_stats: Usage statistics for any tag.
validate_osm_tag: Check if a tag combination is proper.
"Get directions from KL Sentral to Petronas Twin Towers"
"Calculate 30-minute driving isochrone from Grab HQ Singapore"
"Analyze recent mapping activity in Bengaluru"
Run as a standalone server:
npm run dev:httpAccess docs at: http://localhost:8888/api/info
Absolute Paths: Always use full paths in AI tool configurations.
Build First: Ensure you run npm run build after any changes.
Node Version: Use Node.js v18 or higher.
This project is licensed under the MIT License.
OpenStreetMap Contributors
OSRM, Nominatim, Overpass, OSMOSE, Taginfo communities.
Calculate travel time isochrone (areas reachable within time limit) using Valhalla
Inputs: center_longitude (required), center_latitude (required), max_duration_seconds (required), profile, +2 more.
Execute a custom Overpass QL query
Inputs: query (required), timeout, maxsize.
Find amenities near a specific location
Inputs: lat (required), lon (required), radius, amenity_type.
Get detailed information about a specific changeset
Inputs: changeset_id (required), include_discussion.
Get the diff/changes for a specific changeset
Inputs: changeset_id (required).
Calculate distance and duration matrix between multiple points
Inputs: coordinates (required), profile, sources, destinations, +1 more.
Get all OSM elements within a bounding box
Inputs: bbox (required), element_types.
Get detailed route with turn-by-turn directions using OSRM
Inputs: coordinates (required), profile, alternatives, steps, +3 more.
Get detailed information about a specific OSM place
Inputs: osm_type (required), osm_id (required).
Get road/path data for routing between two points
Inputs: start_lat (required), start_lon (required), end_lat (required), end_lon (required), +1 more.
Get usage statistics for OSM tags from Taginfo
Inputs: key (required), include_values, values_limit.
Get OSM tag suggestions and autocomplete based on Taginfo statistics
Inputs: input (required), limit.
Map-match GPS traces to road network
Inputs: coordinates (required), profile, timestamps, radiuses, +1 more.
Solve Traveling Salesman Problem to find optimal route through all points
Inputs: coordinates (required), profile, roundtrip, source, +2 more.
Get detailed information about a specific OSMOSE issue
Inputs: issue_id (required).
Get OSMOSE issues in a specific country
Inputs: country (required), level, item, limit.
Get OSMOSE issues last modified by a specific user
Inputs: username (required), level, bbox, limit.
Get available OSMOSE issue categories/items
Inputs: no named parameters.
Get OSMOSE statistics (issue counts by level, type, etc.)
Inputs: bbox, country.
Search for OSMOSE quality assurance issues with various filters
Inputs: bbox, item, level, country, +3 more.
Get address information from coordinates (reverse geocoding)
Inputs: lat (required), lon (required), zoom.
Search for OSM elements by specific tags
Inputs: tags (required), bbox, around.
Search for changesets with various filters
Inputs: user, display_name, bbox, time, +4 more.
Smart highway search using Taginfo-enhanced queries for major roads, motorways, etc.
Inputs: intent (required), bbox, around, limit.
Search for locations, addresses, or places using Nominatim geocoding service
Inputs: query (required), limit, countrycodes, bounded, +1 more.
Search for Points of Interest (POIs) using Overpass API
Inputs: amenity, shop, cuisine, tourism, +3 more.
Smart POI search using Taginfo-enhanced queries for restaurants, shops, tourism, etc.
Inputs: category (required), subcategory, bbox, around, +1 more.
Search using structured address components
Inputs: street, city, county, state, +3 more.
Snap coordinates to nearest road segments
Inputs: coordinate (required), profile, number.
Validate if an OSM tag combination is commonly used and get suggestions
Inputs: key (required), value (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 30 tools
Most tools have distinct purposes, but there is some overlap that could cause confusion. For example, 'search_pois' and 'search_pois_smart' both search for POIs, and 'get_osrm_route' and 'get_route_data' both handle routing, though their descriptions differentiate them slightly. Overall, the descriptions help clarify boundaries, but agents might need to carefully choose between similar tools.
Tool names follow a highly consistent snake_case pattern with clear verb_noun structures, such as 'calculate_isochrone', 'execute_overpass_query', and 'search_location'. This consistency makes the tool set predictable and easy to navigate, with no deviations in naming conventions.
With 30 tools, the count is on the high side for an MCP server, bordering on heavy. While OpenStreetMap is a broad domain, this many tools might overwhelm agents or indicate potential redundancy. It feels slightly excessive compared to well-scoped servers, but it's not extreme.
The tool set provides comprehensive coverage for OpenStreetMap operations, including geocoding, routing, data querying, quality assurance (OSMOSE), and tag management. It supports CRUD-like actions for OSM elements and changesets, with no obvious gaps that would hinder agent workflows in this domain.
Geocoding, routing, isochrones, traffic, weather, and place search for AI agents. 19 MCP tools.
MCP server for Japan geodata: cadastral lot numbers (chiban) and reverse geocoding, for AI agents.
Geospatial AI MCP server — satellite imagery, embeddings, weather, GNS governance
Routing, geocoding, places, matrix and map tools for location-aware AI workflows.