Railwise for AI
Railwise MCP Server
Railwise data directly in your AI assistant. Connect the MCP server to get live train status, route planning, historical reliability, and connection risk inside Claude, Cursor, and other MCP clients.
At a glance
Coverage
ICE, IC, EC, RE, RB
Access
Remote MCP, no auth
Best for
Route planning and live train checks
How to connect
Add the following URL as a remote MCP server in your application:
https://railwise.eu/api/mcpFor applications that use a JSON configuration file, paste this into your config:
{
"mcpServers": {
"railwise": {
"url": "https://railwise.eu/api/mcp"
}
}
}What the assistant can do
Find the most reliable route from A to B before booking.
Check whether a train is delayed, where it is, and whether the platform changed.
Estimate whether a connection is likely to work based on historical data.
Example prompt
"What is the most reliable route from Frankfurt to Hamburg tomorrow morning?"
Available tools
search_reliable_routes
Find and rank route options by estimated reliability, duration, transfers, and price.
search_station
Search German train stations by name, abbreviation, or partial match and return EVA numbers.
get_train_status
Get live status for a train including delay, platform, current position, and stop sequence.
get_train_reliability
Show historical reliability for a segment with on-time rate, average delay, cancellations, and score.
get_connection_risk
Estimate the probability of making a connection at a station based on historical outcomes.
search_route
Search train connections between two stations and return timing, transfers, and fares.
Notes
Beta
The server is in beta. Tools and behavior may change.
Rate limiting
`search_route` is currently limited to 10 calls per 10 minutes.
Data scope
Covers German long-distance and regional rail.
Authentication
No authentication is required right now. The current beta is free to use.
Important
Railwise is for informational use and does not replace official operator information.
What is MCP?
The Model Context Protocol (MCP) is an open standard that lets AI assistants access external tools and live data. With Railwise connected, answers about German rail can draw on current operational data instead of outdated training data.
Learn more at modelcontextprotocol.io.
Compatible applications
- Claude
- Cursor
- ChatGPT (developer mode only)
- Any MCP-compatible application
FAQ
What does the Railwise MCP Server do?
It gives AI assistants live access to Railwise train data for route planning, train status, historical reliability, and connection risk.
Does it require authentication?
No. The current beta does not require authentication.
Is it connected to the iOS app?
No. The MCP Server is a standalone service for MCP-compatible AI clients.
Standalone service
Built for assistants, powered by Railwise data
The MCP Server is independent from the iOS app, but draws on the same Railwise data: live operations, reliability, and route context.