Know the flight time
before wheels-up.
ML-powered gate-to-gate time estimates for any city pair, sharpened by aircraft type - one call, no model to train or host.
What you get
Gate-to-gate, modeled
Give us a city pair and an aircraft type - get back a realistic block time trained on years of real flights, in a single call.
A realistic block time for any city pair - gate to gate.
Pass route and type, get estimated minutes - taxi included.
R² = 0.9754
An honest input list. A statistical average for the route and aircraft category — not a forecast for one departure time.
- Departure and arrival airport
- Aircraft type, or its category
- Average taxi at both ends
- Wind
- Live en-route weather
Start free - 1,000 requests/month.
Full access to the ML flight-time endpoint on the free tier.
Smarter than a schedule
A model trained on real flights - not a static lookup table - with an honest account of what it does and doesn't factor in.
Gate-to-gate estimate
Returns estimated_minutes - full block time from pushback to arrival gate, including average taxi at both ends.
Aircraft-type aware
Pass an optional aircraft_type (e.g. B77W, A320) for a sharper estimate that reflects real cruise performance.
Trained on real flights
A GradientBoostingRegressor fit on years of historical flight data - R² = 0.9754 against real block times, not a static great-circle guess.
ICAO or IATA inputs
Identify the route with either code style - departure_icao and arrival_icao accept ICAO or IATA airport codes.
One instant call
No model to train, host, or keep current. A single GET returns the prediction in milliseconds - drop it straight into a booking flow or dispatch board.
Transparent methodology
Average taxi time is built into the estimate; winds aloft and en-route weather are not modeled - so you always know what the number does and doesn't include.
Integration
Estimate in one call
Pass a departure, an arrival, and an optional aircraft type - get back gate-to-gate minutes from a model trained on real flights, not a static schedule.
Full API Referenceimport { SkyLink } from "skylink-api";
const sky = new SkyLink({ apiKey: process.env.SKYLINK_API_KEY });
// Gate-to-gate estimate, sharpened by aircraft type
const est = await sky.ml.flightTime({
departure_icao: "EGLL", arrival_icao: "KJFK", aircraft_type: "B77W",
});
console.log(est.estimated_minutes);
// → 437 (7h 17m, incl. avg taxi)Single GET · no model to host · TypeScript types included
* Sample data for illustration only.
Aircraft types
Compare aircraft types
Same route, different metal. Vary aircraft_type to see how block time shifts with cruise performance - useful for fleet planning and schedule padding.
skylink-api — TypeScript SDK · Python SDK
* Sample data for illustration only.
Get a gate-to-gate estimate, sharpened by aircraft type, from a model trained on real flights - in one call.
Any route, a real number
Gate-to-gate minutes for any city pair - taxi included, aircraft-type aware.
Trained, not guessed
A regression model fit on years of real flights - with an honest scope.
Build your query
Prefer fine-grained control?
Set the route and aircraft type yourself and read estimated_minutes straight off the response.
Ready to predict?
Gate-to-gate estimates, aircraft-type aware, from a model trained on real flights - all behind one API key. Drop it into booking flows, dispatch boards, and EFBs.