Skip to main content

db-fahrplan-mcp

PyPI Python CI bahn.de canary License: MIT Glama score

Live Deutsche Bahn timetables for Claude, Cursor, and any MCP client: connections, every fare option, cheapest time of day, departure boards, delays, platform changes, all stops of a train, coach sequence, stations near you.

It talks directly to the JSON endpoints behind bahn.de — the same ones the website uses. No API key. No proxy. No third-party mirror to go down. You get what bahn.de shows, including real-time data.

uvx db-fahrplan-mcp

What it can answer

Question Tool
"Krefeld to Köln Messe, there by 8 on Sunday" db_journeys (arrive_by)
"…with my BahnCard 50, two kids, bike, max one change, via Düsseldorf" db_journeys (bahncard, children_ages, bike, max_transfers, via)
"Only Deutschlandticket trains" db_journeys (deutschlandticket)
"What tickets exist for that 09:35, and what do they cost?" db_journey_offers — Super Sparpreis / Sparpreis / Flexpreis / regional day tickets
"When is it cheapest to go to Berlin on the 12th?" db_best_price — per time band of the day
"What leaves Köln Hbf in the next 30 min?" db_departures / db_arrivals — live, with delay, platform change, cancellation
"Where is ICE 555 right now? Does it still stop at Hagen?" db_trip_details — every stop, live
"Which sector do I stand in for 1st class on ICE 109?" db_train_formation — coach sequence and platform sectors
"Stations near 50.94, 6.96" db_nearby
"Is 'Frankfurt Flughafen' the Fernbahnhof or Regionalbahnhof?" db_search_station

Every journey carries a recon_token (→ fares), every leg and board row a journey_id (→ stops). Times are Europe/Berlin wherever the server runs; (+7) means 7 minutes late; (+1d) means after midnight. Occupancy (load), amenities (WiFi, bike, step-free, quiet zone…), real-time platforms and disruption notes are included.

Install

Claude Code

claude mcp add db-fahrplan -s user -- uvx db-fahrplan-mcp

Claude Desktop / Cursor / any MCP clientclaude_desktop_config.json:

{
  "mcpServers": {
    "db-fahrplan": {
      "command": "uvx",
      "args": ["db-fahrplan-mcp"]
    }
  }
}

No uv? pipx install db-fahrplan-mcp or pip install db-fahrplan-mcp, then use "command": "db-fahrplan-mcp".

Remote / HTTP

db-fahrplan-mcp --transport streamable-http --host 127.0.0.1 --port 8000

Binds to localhost by default. Put it behind a reverse proxy with auth before exposing it.

Example

Krefeld Hbf → Köln Messe/Deutz, arrive by 08:00 Sunday

db_journeys("Krefeld Hbf", "Köln Messe/Deutz", "2026-08-30 08:00", arrive_by=true, format="text")

Krefeld Hbf → Köln Messe/Deutz, arrive by 2026-08-30 08:00 (Europe/Berlin)

06:35 → 07:23 · 48 min · 0× change · 25.80 €
   RE7       Krefeld Hbf 06:35 Gl.5 → Köln Messe/Deutz 07:23 Gl.1

…and on a bad day (real output, signal-box failure at Montabaur):

15:53 → 17:09 · 76 min · 0× change
   ICE 125   Köln Hbf 15:53 Gl.2 A-C → Frankfurt(Main)Hbf 17:09 Gl.1 CANCELLED
      ! Ein defektes Stellwerk im Raum Montabaur beeinträchtigt den Bahnverkehr …
   ! Verbindung fällt aus

format="text" is compact and cheap in tokens; format="json" (default) is structured.

Configuration

Env var Default
DB_FAHRPLAN_RATE_PER_MIN 30 Token-bucket limit on requests to bahn.de
DB_FAHRPLAN_TIMEOUT 30 HTTP timeout, seconds
DB_FAHRPLAN_LOG WARNING Log level (stderr)
DB_FAHRPLAN_BASE_URL https://www.bahn.de/web/api For testing against a mock

How it compares

db-fahrplan-mcp DB API Marketplace servers¹ transport.rest wrappers²
API key none client id + secret none (shared 100 req/min)
Journey planning ❌ (timetables only)
Prices, fare breakdown, best-price day
BahnCard / children / via / bike / max changes partial
Occupancy, amenities, coach sequence
Real-time delays, platform changes, cancellations
Runs offline tests, weekly live canary

¹ e.g. PaulvonBerg/db-mcp-server, jorekai/db-timetable-mcp. ² e.g. AnythingMCP's DB connector.

Development

git clone https://github.com/capraCoder/db-fahrplan-mcp && cd db-fahrplan-mcp
python -m pip install -e ".[dev]"
pytest              # offline, replays recorded bahn.de responses (tests/fixtures)
pytest -m live      # hits bahn.de — Krefeld→Garmisch multi-modal trip, fares, best price, boards
ruff check . && mypy

CI runs on Linux + Windows, Python 3.10–3.13, on both MCP SDK 1.x and 2.x. A weekly canary runs the live tests and opens an issue if bahn.de changes something.

Legal, honestly

Full text: DISCLAIMER.md. The short version:

  • Unofficial. Not affiliated with, endorsed by, or supported by Deutsche Bahn AG. "Deutsche Bahn" and "DB" are trademarks of Deutsche Bahn AG. The software ships no DB data and performs no systematic extraction — each query fetches only what the user asked for.
  • Undocumented endpoints. bahn.de can change or block them at any time. The canary will notice within a week; please open an issue with the raw response if you hit it first.
  • bahn.de's terms of use restrict automated access. This is a personal, non-commercial tool with an honest User-Agent and a rate limiter (30 req/min by default). Use it for yourself; do not build a scraping farm on it. Cloud/VPN IPs are sometimes blocked (HTTP 403).
  • Prices are informational, as bahn.de quotes them for the given travellers at that moment. Booking happens on bahn.de.
  • Timetable changes (mid-December, mid-June) can make far-future queries temporarily odd.
  • No liability for missed connections or wrong fares — see DISCLAIMER.md.

Citing

If this server is part of published work, cite it (metadata in CITATION.cff; GitHub's "Cite this repository" button renders it):

capraCoder (2026). db-fahrplan-mcp: Deutsche Bahn timetables as an MCP server (Version 1.0.0) [Software]. https://github.com/capraCoder/db-fahrplan-mcp

@software{capracoder_db_fahrplan_mcp_2026,
  author  = {{capraCoder}},
  title   = {db-fahrplan-mcp: Deutsche Bahn timetables as an MCP server},
  version = {0.2.1},
  year    = {2026},
  url     = {https://github.com/capraCoder/db-fahrplan-mcp}
}

Release artifacts carry OpenTimestamps proofs (*.ots, Bitcoin-anchored): existence and integrity are provable without any third-party archive.

Licence

MIT.

Release files for db-fahrplan-mcp 1.0.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for db-fahrplan-mcp 1.0.0
File Size Uploaded
db_fahrplan_mcp-1.0.0.tar.gz 22.7 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for db-fahrplan-mcp 1.0.0
File Interpreter ABI Platform
db_fahrplan_mcp-1.0.0-py3-none-any.whl Python 3 none any Details

Total release size: 42.7 kB

Release files / db_fahrplan_mcp-1.0.0.tar.gz

Download URL db_fahrplan_mcp-1.0.0.tar.gz
Size 22.7 kB
Tags Source
SHA-256 checksum
How to use checksums
6adebda1243572319e103e93f222d7d73916ab295632308852c12f8d83f93fae
BLAKE2b-256 checksum
How to use checksums
cd16b81a988a24fdff3e22708272309f9e7ed6cfd07dc89ebbb922a5c29af983
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 31, 2026.

Transparency log

Release files / db_fahrplan_mcp-1.0.0-py3-none-any.whl

Download URL db_fahrplan_mcp-1.0.0-py3-none-any.whl
Size 20.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
9ecdc8e02705285e5354103700355b5b70919df0b814aea04a1941c46c837138
BLAKE2b-256 checksum
How to use checksums
eaabfad70d8d1a52cf6ea564b44ff1fcacf44b7137395f1085e4f0b20439d12e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 31, 2026.

Transparency log

Release history Release notifications | RSS feed

This release

1.0.0 This release

2 release files

0.2.2

2 release files

0.2.1

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page