Skip to content
SchedulesMCP

Point-to-point sailing schedules, compared the way a desk actually books them.

Point-to-point sailing schedules across 5,000+ lanes and 255 ports, grouped by the physical vessel, so every carrier selling slots on the same hull shows up on one card. Each sailing carries transit time, transshipment legs, cut-offs and a reliability score built from arrivals we have actually seen.

MCP call

schedulesmcp.com · live
{ "tool": "get_schedules",
"origin": "CNSHA",
"destination": "NLRTM",
"from": "2026-11-02" }
response
{ "carrier_code": "MAEU", // first of sailings[]
"vessel_name": "MAERSK HANGZHOU",
"voyage_number": "445W",
"published_departure": "2026-11-04T18:00:00", // port local time
"published_arrival": "2026-12-06T08:00:00",
"transit_days": 32,
"transship_via": ["SGSIN"],
"cy_cutoff": "2026-11-02T12:00:00" }
get_schedulesfind_best_sailingscompare_lanes
5,000+ lanes · 255 ports72,000+ forward sailingsReliability from observed arrivalsMCP tools + REST API

What it does

Sailing schedules built for the decisions you make at the desk.

Vessel-first sailing cards

One card per physical hull. When three carriers sell slots on the same vessel, you see one sailing with three ways to book it, not three near-identical rows.

Reliability scored from real arrivals

On-time performance per lane comes from arrivals we observed through the tracking layer, not a carrier’s own marketing number. Thin lanes are honestly labelled early-data, never faked.

The detail a desk books on

Transit time, transshipment ports and legs, cut-offs, and emissions on every option: the things that decide a booking, on the card.

Type a port, not a code

Search “Shanghai”, not CNSHA. Names resolve to the codes our data actually uses, so you find the lane on the first try.

5,000+ lanes, refreshed weekly

Published schedules cached across 5,000+ lanes and 255 ports, enriched against live vessel movements, with 72,000+ forward sailings.

Call it as tools

Point-to-point schedules as MCP tools or REST endpoints.

The same functions an assistant can call in plain language, your stack can hit over REST, behind one key, with webhooks to push the changes.

Read the SchedulesMCP docs
SchedulesMCP · tools
  • get_schedules Published sailings on a lane: vessel, voyage, ETD/ETA, cut-offs.
  • find_best_sailings Best upcoming sailing per carrier on a lane, ranked, with filters.
  • compare_lanes Rank carriers on one lane by on-time reliability and transit.
  • carrier_reliability On-time % and delay distribution for a carrier, by SCAC.
  • carrier_league Carrier on-time league table.
  • watch_lane Schedule alerts on a lane: blank sailings, cut-offs, slips (paid).
  • port_congestion Congestion level and expected added wait at a port.
  • active_cyclones Live tropical cyclones on ocean trade lanes.
  • lane_storm_exposure Active-cyclone exposure along a lane.

Questions

The things people ask first.

From arrivals we actually observed through the tracking layer, per lane, with the sample size shown. Thin lanes are honestly labelled early-data, never inflated.

Because several carriers often sell slots on the same hull. One vessel-first card shows one sailing and every way to book it, instead of three near-identical rows.

Schedules for 5,000+ lanes across 255 ports, refreshed weekly and enriched against live vessel movements, with 72,000+ forward sailings.

Yes. Type “Rotterdam”, not a UN/LOCODE. Names resolve to the codes our data actually uses.

Make your first SchedulesMCP call today.

Free tier, five-minute first call, no front-end of ours required.