- DATA
- MAPS
- LOGIC
- HUMAN
Short answer
How do I find locations in a bounding box with my AI?
Give your AI a set of coordinates defining a rectangle and tell it what kind of places you're looking for. The AI uses the Pelias Geocoder to scan that area and return a list of matching points of interest. You'll get a clean list of names and locations based on your exact boundaries.
The results
What your AI delivers
The AI processes your spatial requests into structured data.
DATA
Filtered location lists
The AI identifies specific points of interest within your defined rectangle. You get names, addresses, and coordinates for everything that fits.
MAPS
Spatial boundaries
Your AI respects the exact rectangular limits you set. It won't pull data from outside the box you defined.
LOGIC
Layered filtering
You can narrow down results by specific categories. The AI filters the search to show only the types of places you actually care about.
HUMAN
Manual verification
The AI provides the raw data, but you decide which locations are actually relevant to your project. You handle the final selection and decision making.
The workflow
What your AI does when the request arrives.
The AI translates your plain English descriptions into precise geographic queries.
Define the area
The AI takes your description of a neighborhood or a set of coordinates and builds a rectangular search area.
search_bounding_boxApply filters
If you ask for only coffee shops or parks, the AI narrows the search to those specific layers.
search_layer_filterExecute search
The AI runs the query against the Pelias database to find all matching points within your box.
search_bounding_boxFormat results
The AI organizes the found locations into a readable list or table for you to use.
search_layer_filter
Try it
Copy these to start.
Use these prompts to trigger the geocoding tools.
Starting points
Swap out the bracketed text with your actual coordinates or city names.
Pelias Geocoder Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_9xauhtmLx4FJnpssXKIJfawJuO9h0Fd6rFMjRoZ7/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Pelias Geocoder capabilities are ready to use.
{
"mcpServers": {
"pelias-geocoder-mcp": {
"url": "https://edge.vinkius.com/vk_preview_9xauhtmLx4FJnpssXKIJfawJuO9h0Fd6rFMjRoZ7/mcp"
}
}
}Find all restaurants within this bounding box: [insert coordinates].
List all parks located inside the rectangle defined by [lat1, lon1] and [lat2, lon2].
Search for gas stations in the area bounded by these coordinates: [insert coordinates].
Show me all libraries within this specific geographic box: [insert coordinates].
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Pelias Geocoder once, then ask.
Click the link to connect the MCP to your client. Your credentials stay encrypted, and you can immediately start asking your AI for location data.
Connect Pelias Geocoder to your AIFAQ
How this task behaves
- 01
Can the AI add new locations to the map?
No. The AI can only search and retrieve existing data from the Pelias database.
- 02
Can the AI delete locations it finds?
No. The AI has no permission to modify or delete any data in the geocoder.
- 03
Do I need to provide exact coordinates?
It helps. While you can describe an area, providing specific bounding box coordinates ensures the most accurate results.
- 04
Can I filter by specific business types?
Yes. You can ask the AI to only look for specific layers, like restaurants or transit stops.
- 05
Will the AI show me the results on a visual map?
The AI returns the data as text or a list. You will need to paste those coordinates into a mapping tool if you want to see them visually.
More questions about Pelias Geocoder? The Connector page answers them. See everything the Pelias Geocoder Connector can do
Connect Pelias Geocoder to Claude, Cursor, ChatGPT & more
