# Mapbox MCP for AI Agents AI Agent Connect > Mapbox MCP lets you bring high-quality maps, routing data, and geospatial analytics into your AI agent. It handles everything from generating static map images and calculating complex travel matrices to managing your Mapbox datasets and publishing custom tilesets. ## Overview - **Category:** developer-tools - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_feuCsMrdUMQVcySXmvWHZ3qiBeFWtDMTgS8QvxFG/ai-agent-connect - **Tags:** mapping, geospatial, navigation, gis, location-services ## Description This Mapbox MCP lets you bring high-quality maps, routing data, and geospatial analytics directly into your conversation with your AI agent. You shouldn't have to jump between tabs and manually copy data just to see how long a delivery route takes or to grab a map image for a report. Instead of hunting for coordinates or building out a travel matrix in a separate dashboard, you just ask your agent to do it. It handles the heavy lifting of fetching raster and vector tiles, generating static images, and pulling specific features from your datasets. If you manage a fleet or build location-based apps, you can manage your tileset sources and tokens without leaving your workspace. It turns your agent into a functional geospatial assistant that understands distance, time, and map data in real-time. You can find this and other tools in the Vinkius catalog to get your setup running quickly. ## Tools ### list_tokens View all active access tokens for your account. Use this to audit who has access to your Mapbox data. ### publish_tileset Make a Mapbox Tiling Service tileset live for your use. Use this to push new map layers to production. ### searchbox_forward Perform a forward search using the Search Box API. This helps your agent find locations based on text input. ### searchbox_retrieve Get full details for a specific searchbox suggestion. Use this to get more context on a suggested location. ### searchbox_reverse Perform a reverse search using the Search Box API. Use this to find locations based on coordinates. ### searchbox_suggest Get autocomplete suggestions for location searches. This helps your agent understand what locations a user might mean. ### geocode_batch Perform batch geocoding for multiple addresses at once. This saves time when processing large lists of locations. ### geocode_forward Convert a specific address into geographic coordinates. Use this to get exact points for your maps. ### get_dataset_feature Pull a specific feature from one of your Mapbox datasets. It lets your agent see specific data points in your collection. ### get_isochrone Calculate the area reachable within a set time or distance. This is great for seeing how far a delivery can go. ### get_map_matching Snap fuzzy GPS coordinates onto the actual road network. This ensures your agent understands real-world movement. ### get_static_image Generate a static map image for a specific view. Use this to create snapshots for reports or presentations. ### get_tilequery Get data about specific features at a specific coordinate. This lets your agent know exactly what is at a point. ### get_vector_tile Fetch a vector tile from a Mapbox-hosted tileset. Use this for high-performance, data-rich map layers. ### list_datasets See all the datasets associated with your Mapbox account. This helps your agent see what data is available to query. ### create_tileset_source Create a Mapbox Tiling Service source for your map layers. This helps you organize custom map data. ### create_token Generate a new access token for your Mapbox account. Use this to manage permissions for your agent. ### delete_token Remove an existing access token from your account. This keeps your account secure by removing old keys. ### geocode_reverse Turn geographic coordinates into a readable street address. This is perfect for identifying locations from GPS pings. ### get_matrix Get travel times or distances between many different points. Use this to build a full travel matrix for a fleet. ### get_directions Get turn-by-turn instructions and optimal routes between locations. This provides the actual path for a driver. ### get_feedback Retrieve user feedback from your Mapbox interactions. Use this to see how people are interacting with your maps. ### get_optimization Grab a previously submitted optimization solution for your route. This helps you compare different routing plans. ### submit_feedback Send user feedback back to the Mapbox system. Use this to report issues or provide data for map improvements. ### submit_optimization Solve the Traveling Salesperson Problem for complex fleet routing. This finds the most efficient way to hit multiple stops. ### get_raster_tile Fetch a raster tile image from a Mapbox-hosted tileset. This is ideal for getting high-quality map imagery. ### update_dataset_feature Change the details of a specific feature in a dataset. Use this to keep your geographic data up to date. ## Prompt Examples **Prompt:** ``` How long will it take to drive from Chicago to Nashville with 3 stops in between? ``` **Response:** ``` I've calculated the optimal route for your trip from Chicago to Nashville with 3 stops. Here is the breakdown: | Leg | From | To | Est. Time | Distance | |---|---|---|---|---| | 1 | Chicago | Stop A | 2h 15m | 130 miles | | 2 | Stop A | Stop B | 1h 45m | 85 miles | | 3 | Stop B | Nashville | 3h 10m | 210 miles | **Total Travel Time:** 7h 10m **Total Distance:** 425 miles ``` **Prompt:** ``` Can you show me a map of the downtown area in Austin? ``` **Response:** ``` I've generated a static map image of downtown Austin for you. You can view it here: [Map Image URL]. This view shows the central business district and nearby parks. ``` **Prompt:** ``` What's located at these coordinates: 34.0522, -118.2437? ``` **Response:** ``` At those coordinates (34.0522, -118.2437), the following features were identified: * **Primary Feature:** Los Angeles City Hall * **Area Type:** Urban / Government District * **Nearby Landmarks:** Grand Park, Little Tokyo Would you like me to find more details about any of these specific features? ``` ## Capabilities ### Generate static map images Create a custom map image for a specific coordinate and zoom level. ### Calculate travel matrices Get travel times and distances between many different points at once. ### Fetch map tiles Retrieve both vector and raster tiles from Mapbox-hosted tilesets. ### Query dataset features Pull specific data points or features from your existing Mapbox datasets. ### Manage tileset sources Create, update, and publish custom tileset sources for your maps. ### Perform geocoding Convert addresses to coordinates or find street names for specific points. ## Use Cases ### Optimizing complex delivery routes A logistics manager wants to find the fastest way to hit 10 stops. They ask the agent to solve the route, and it uses submit_optimization. ### Generating map snapshots for reports A developer needs to see a map of a specific park. They ask for a static image, and the agent uses get_static_image. ### Identifying features at specific coordinates A researcher needs to know what's located at a specific coordinate. They ask the agent, which then uses get_tilequery to describe the features. ### Updating geographic dataset records A data analyst needs to update a city's park location. They ask the agent to modify the feature using update_dataset_feature. ## Benefits - Get instant travel times and distances between multiple stops using get_matrix to plan logistics without manual math. - Create and publish custom map layers quickly with create_tileset_source and publish_tileset to keep your data live. - Turn messy GPS data into clean road paths using get_map_matching to ensure your agent understands real-world movement. - Fetch specific geographic data points on the fly with get_tilequery to populate your reports with real-time details. - Automate address lookups for large lists of locations using geocode_batch to save hours of manual data entry. ## How It Works The bottom line is you get a direct line to Mapbox's mapping engine inside your favorite AI client. 1. Subscribe to the Mapbox MCP through the Vinkius marketplace. 2. Add your Mapbox Access Token to your AI client configuration. 3. Ask your agent to calculate routes, generate maps, or query geospatial data. ## Frequently Asked Questions **Can the Mapbox MCP help with fleet routing?** Yes, it can. It uses the Mapbox engine to solve complex routing problems, find the best order for multiple stops, and provide total travel times for your fleet. **Does this Connector work for reverse geocoding?** Yes, it can. You can give your agent coordinates, and it will return the readable street address and location details for you. **Can I manage my Mapbox datasets with this?** Yes, you can. The Connector allows your agent to list your datasets, retrieve specific features, and even update those features using natural language. **How does this help with travel matrices?** It allows your agent to calculate the distance and travel time between many different points simultaneously, which is great for planning multi-stop journeys. **Can I get static map images?** Yes, you can. You can ask your agent to generate a static map image for any coordinate, and it will provide a URL for you to view or share. **Does this Connector support batch geocoding?** Yes, it does. You can provide a list of addresses, and the agent will process them all at once to give you the corresponding geographic coordinates. **Can I calculate driving directions between two points using this server?** Yes! Use the `get_directions` tool. You need to provide a routing profile (like `mapbox/driving`) and the coordinates as a semicolon-separated string of longitude and latitude. **How do I find information about map features at a specific coordinate?** You can use the `get_tilequery` tool. By providing a tileset ID and the longitude/latitude, the agent will retrieve data about the specific features located at that point. **Is it possible to update my Mapbox datasets through the AI?** Yes, the `update_dataset_feature` tool allows you to modify a specific feature within a dataset by providing the dataset ID, feature ID, and the new GeoJSON data.