Freight MCP server for AI assistants and agents.
ForTeams building copilots and agents that need to answer freight questions
An assistant is only as good as the tools it can call. Ask most freight copilots “where is this container?” and they improvise, because they have nothing real to reach for. Navo24 is the thing to reach for.
A freight MCP server for every product
Each Navo24 product runs its own Model Context Protocol server. Connect your AI assistant once, and it can:
- Track:
get_container_detail,get_demurrage_free_time,get_port_congestion. - Schedule:
get_schedules,find_best_sailings,compare_lanes. - Load:
plan_load,suggest_containers,export_plan.
The model picks the tool, calls it and answers from the result. ETAs and free time come from the carrier’s own data, so the assistant has nothing to make up.
Grounded answers, by construction
The tools return structured, sourced data, and the assistant’s answer keeps that grounding: a status backed by a real carrier event, a reliability score with its sample size, a load plan with its compliance flags. The hard part of a trustworthy freight copilot is the data layer. That is the part Navo24 builds, with tracking across 240+ carriers.
Connect your assistant in minutes
Add the server URL to Claude Desktop, Cursor or any other MCP client.Pass your API key as a bearer header.Ask about a container and the model calls the tracking tool on its own.
The MCP quickstart has the endpoint and a config snippet for each server.
Take one tool or all three
Wire in the tracking tool today and add schedules and loading when you need them. It is the same key and the same protocol, with no platform to adopt. An MCP tool call and a REST call hit the same backend, so your assistant and your own code see the same numbers.
Questions about a freight MCP server
What is a freight MCP server? A server that exposes freight data as Model Context Protocol tools. An AI assistant calls them directly for container status, sailings or a load plan instead of guessing from its training data.
Which AI assistants can use the Navo24 tools? Any MCP client, such as Claude Desktop, Cursor or an agent you build yourself. You add the server endpoint and pass your API key as a bearer header.
Do the MCP tools return the same data as the REST API? Yes. An MCP tool call and a REST call hit the same backend behind the same key. Use MCP where a model drives the work and REST where your own code does.
Do I have to connect all three tools? No. Start with tracking and add schedules and load planning later with the same key.