Free tool · no account
Drop in your GPX, give your flat pace, and you get your estimated finishing time and a split time for every kilometre. Elevation is integrated gradient by gradient, not by a flat rule of thumb.
Drop your file .gpxor click to browse
Drop in a track to get your estimated finishing time and your split times kilometre by kilometre.
Other calculators ask for distance, vertical gain and pace. Three numbers cannot tell a steady col apart from a sawtooth profile with the same gain, even though the second costs considerably more. Here the track is read point by point.
No "+10 min per 100 m". Every gradient gets its measured energy cost, and the sum is your finishing time. The effort coefficient is shown for each kilometre.
A descent does not give back the time a climb took, and a very steep one costs you time. Most calculators treat elevation loss as a linear bonus, or ignore it entirely.
Once the finishing time is set, the same race clock places your gels and your sips to the minute and the kilometre. That is the natural next step, and it is what the rest of the site does.
The estimated time is a projection, not a promise: it is worth exactly what the reference pace you entered is worth. Its value is in telling you how far the terrain moves you away from a per-kilometre calculation.
Flat-equivalent distance is the number to remember. It explains why a 21 km mountain race is prepared for like a very long half marathon, and why you need to fuel it accordingly.
The split times assume you hold the same intensity from start to finish. In a race almost nobody does: use them as a consistency reference, not a contract.
You have to convert terrain into time. The common method adds a flat allowance per band of vertical gain, often 10 minutes per 100 m, on top of your road pace. That is quick but crude: it ignores the gradient at which the climbing happens, and the same 1 000 m spread over 10 km or packed into 2 km do not cost remotely the same. This calculator works differently: every pair of GPX points gets an effort coefficient from its gradient, the whole lot is summed into a flat-equivalent distance, and your reference pace is applied to that.
It is the distance that would cost you the same effort on flat ground. A 21 km course with 1 050 m of gain works out at roughly 27 km of flat: at 5:30/km on the flat that is about 2 h 27, not 1 h 56. That number carries the whole calculation, and it is shown so you can check it.
The pace you could hold on flat ground for the whole duration of the race, fatigue included, not your 10 km pace. For a four-hour trail race, use your long-run pace rather than your race pace. The model translates terrain; it does not simulate you fading.
Only indirectly. The model sees gradients, not roots, scree or mud. On very technical ground, pad your reference pace. It does account for steep descents slowing you down though: the coefficient bottoms out around −12 % then rises, and past −35 % you are slower than on the flat.
Drop your track in, pick your pace or your target time, and the table fills in: split time, duration, pace and elevation for every kilometre. It downloads as CSV, ready to print or tape to a wristband.
From the energy-cost polynomials of Minetti et al. (2002), which measure the cost of running and walking by gradient, with three corrections specific to trail running: the saving from running downhill is only partly converted into speed, braking costs you beyond −10 %, and the model switches to power hiking on steep gradients. The constants are documented and adjustable.
The same race clock places your carbohydrate and your hydration: what to take, at what minute, at what kilometre. First race free.
Add my nutritionSee also the nutrition calculator and ultra running nutrition.