Recorded tracks generally tend to overestimate elevation gain because GPS/GNSS altitude measurements are not very accurate. A recorded track is essentially a long sequence of more or less inaccurate elevation values, and these errors accumulate when calculating total elevation gain. Applications try to filter them out, but sometimes the measurements fluctuate so much that filtering cannot remove all the errors. The result is an overestimated elevation gain. However, the opposite can also happen.

Planned routes often provide a more reliable estimate because they use elevation data from a digital terrain model, which generally produces a smoother elevation profile. However, these data also require filtering, which can lead to underestimated elevation gain. Again, this is not a universal rule. Planned routes can also overestimate elevation gain, for example in mountainous areas where a road follows a river through tunnels and over bridges, while the terrain model returns elevations below the bridges or above the tunnels.

Here on the Mapy.com forum, as well as in discussions elsewhere, users have repeatedly reported that Mapy.com tends to underestimate elevation gain for planned routes. However, these differences are often relatively small.

For your route (900m planned and 1300m recorded), I would roughly estimate the actual elevation gain to be somewhere in between, probably closer to the planned value, perhaps around 1000m. This is only an illustrative estimate, of course, without examining the actual route. In that case, the planned route would underestimate elevation gain by about 10%, while the recording would overestimate it by about 30%. Such differences are quite common, and all mapping applications struggle with them. None is perfect.

Mapy.com is aware that its planned elevation gain can sometimes be significantly underestimated, especially on very long routes. The developers have acknowledged this issue and indicated that they intend to address it. However, the details of their algorithms are internal information, just as with other mapping services.

I would like to emphasize once again that differences in elevation gain of several tens of percent are not unusual and are certainly not a problem unique to Mapy.com.

Regarding your second example, I'm not sure exactly what you're comparing. If you mean the difference between the minimum and maximum elevations shown in the elevation profile of a planned route, then the calculated total elevation gain should indeed not be smaller than that difference. Similar discrepancies can occur in other route planners as well, not just Mapy.com. However, Mapy.com can sometimes underestimate elevation gain quite significantly. The Mapy.com team is aware of this issue and is gradually working to improve the calculations. However, they do not publicly share the details, as I mentioned above.

I am currently working on an educational article explaining this rather complex subject in detail. I expect to finish it later this year (2026). Once it is ready, I will post a link here.

Here is a list of threads on this forum discussing these issues.

Výškový profil Stopař (CZ) – GPS and filtering
Difference in elevation gain between routes – why? (EN) – 9301m vs. 19237m
Höhenmeter extrem zu viel (DE) – 40m vs. 1800m
Rozdiel v nastúpaných / naklesaných metroch (SK) – 718m, 585m, 620m
Převýšení delších tras. (CZ) – 9500m vs 25000m
Zu geringe Höhendifferenzen im Höhenprofil? (DE) – Mapy, Swisstopo, Komoot
Improve route planning + elevation calc on routes with tunnels (EN) – 1400m vs. 500m
Mapy underestimates elevation (EN) – underestimated by 20–25%
Dislivelli errati in eccesso (IT) – over 2000m vs. 1390m
Elevation gain issues (EN) – planned route 900m, recorded track 1300m