How fast is Loopa? We are measuring it.
Every app says it is fast. Almost none of them will tell you what they measured, on what, or how many times. This page is where we publish ours, release after release.
848 logging events from 24 members, measured between 20 and 31 August 2026. No figure is published below yet. Here is exactly how close they are.
The measurement
From deciding to log, to seeing it
The clock starts the moment you commit to logging something: tapping the camera, opening the barcode scanner, reaching for a saved meal, typing the first letter of a search. It stops when the entry is on your screen.
That is the demanding definition, and it is the one that matches your experience. The easy thing to publish is how quickly our server answers, which has almost nothing to do with how long logging a meal actually takes you. We measure the round trip separately and it is not part of this figure.
Failed logs count too. A build that fails quickly is not a build that logs quickly, so every attempt stays in the sample.
The standard we hold ourselves to
Nine times in ten, not on average. An average hides the bad days; the ninetieth percentile is the slow log you actually remember. The 500-event floor is what makes a percentile evidence rather than arithmetic.
Where the count stands
848 logs in, and holding the line
Eleven days of measurement, 24 members, every log counted. The two figures this page exists to publish are the two we use most carefully, and neither has reached its floor yet.
| Figure | Target | Events so far | Publishes at |
|---|---|---|---|
| Repeating a meal | under 10s, 9 times in 10 | 41 | 500 |
| Photographing a plate | under 20s, 9 times in 10 | 31 | 500 |
| Every logging method, combined | measured, not targeted | 848 | . |
We could publish a percentile from forty events today. It would be a number, it would probably look good, and it would not be evidence, which is the one thing this page is for. The floor stays where it is, and the figures publish when they clear it.
What a measurement contains
Which method, how long, whether it worked
A sample is four things: the logging method you used, the number of milliseconds, whether it succeeded, and your device model. That is the whole record.
It carries nothing about what you ate, what you weigh, or anything you typed. There is nowhere in the record to put those, so nothing can arrive there by accident. It is first-party and it stays with us. Loopa uses no advertising or analytics company, so there is nowhere else for it to go.
Your own samples are included in your data export, and they are deleted with your account.
Questions
The obvious ones
Why publish this at all?
Because “feels fast” is not a claim anyone can check, and every app says it. A number with a sample count and a date beside it is checkable, and this is where ours goes, release after release.
What exactly is being timed?
The gap between deciding to log something and seeing it in your diary. The clock starts when you tap the camera, open the scanner, reach for a saved meal or type the first letter of a search. Not when the app gets round to calling our server. It stops when the entry is on your screen.
Why not just time the server?
Because server time is not the time you spend. Our server answering in 200 milliseconds is no comfort to someone who spent nine seconds getting to the right screen. We measure the server separately and it is not part of this figure.
Does this send anything about my food or my weight?
No. A sample is: which method you used, how many milliseconds it took, whether it worked, and your device model. There is no field for anything else, so there is no route for it to travel. It is first-party, and it does not go to an advertising or analytics company, because we do not use any.
Can I turn it off?
There is nothing here to turn off in the sense that matters. No profile is built and nothing is shared. If you would rather it stopped anyway, ask us and we will tell you exactly how, or turn it off for your account ourselves.
What do you publish after each release?
The current figure, on this page, measured the same way every time. A performance page that only ever moves one way is an advertisement, not a measurement.
Why is there no figure on this page yet?
Because neither measurement has reached 500 events yet. As of 31 August 2026 we have 848 logging events from 24 members, but they are spread across every way of logging: repeating a meal accounts for 41 of them and photographing a plate for 31. A ninetieth percentile drawn from forty events moves several seconds on a single slow log, so publishing one would tell you about the sample rather than about the app. The sample is still small, so this takes as long as it takes.
Try it and judge for yourself
Loopa Basic is free, with no card and no time limit. Numbers or no numbers, your own thumb is the instrument that matters.