GuidesDiscussions
Guides
  1. HERE Location Reasoning Docs

Overview

HERE Location Reasoning

HERE Location Reasoning (HLR) is a HERE-powered agentic AI capability that enables LLMs, AI agents, enterprise systems, and in-car assistants to make better decisions using location intelligence. It exposes HERE's location services as reasoning-ready tools via the Model Context Protocol (MCP), turning raw REST APIs into capabilities that AI agents can discover, invoke, and chain together autonomously.

What is HLR?

Location Reasoning sits between AI/agent clients and HERE's location APIs, handling tool orchestration, state management, and API calls so agents can request complex location operations — like "find an optimal delivery route with traffic and EV charging constraints" — and get grounded, API-backed results rather than pure LLM guesses.

MCP is model-agnostic: any LLM from any provider can use HLR as long as it supports tool calling. There is no dependency on a specific model vendor.

Tool Categories

HLR provides location tools across the following categories:

  • Geocoding — Convert text addresses or place names into geographic coordinates, and coordinates back into human-readable addresses.
  • Routing — Calculate optimized routes between locations with distance, duration, turn-by-turn instructions, and support for EV charging and rest stops.
  • Search & Discovery — Find places of interest (POIs, EV stations, services) near a location or along a route corridor.
  • Traffic — Get real-time traffic flow data for locations, corridors, or route segments.
  • Isolines — Calculate reachable areas from a point within time or distance limits (drive-time polygons).
  • Map Rendering — Generate interactive map visualizations with support for routes, traffic, and isolines.

Architecture

HLR bridges three layers:

graph TD
    A[AI / Agent Layer] -->|MCP protocol| B[Reasoning / Orchestration Layer]
    B -->|HERE API calls| C[HERE Location Intelligence Layer]

    A --- A1[LLMs, Copilots, Voice Assistants, In-Car Assistants]
    B --- B1[MCP Server, Tool Routing, Session Management]
    C --- C1[Maps, Routing, Traffic, Geocoding, POIs, EV, Transit]
  1. AI / Agent Layer — Your LLM, copilot, or agent framework that sends natural-language-driven tool calls.
  2. Reasoning / Orchestration Layer — The HLR MCP server that receives tool calls, manages session state, and coordinates execution.
  3. HERE Location Intelligence Layer — HERE's production APIs that provide the underlying location data and computation.

Data flows downward (agent requests tool execution) and responses flow upward (location results returned to the agent).

Deployment

The HLR MCP server is available at:

https://hlr.ai.here.com/mcp

Authentication, request routing, and credential management are handled for you — send MCP requests with a HERE bearer token and receive location results.

Next Steps

Ready to make your first request? Head to the Getting Started guide.