Katoptron shows the score, the formula, and the error margin. Kato reads your check-in and picks one move for the next 2 hours. 7-day trial.
7 days, card required, nothing charged before day 7. Built solo · local compute · cancel in one click.
Eight hours of sleep, low stress, decent hydration. The model puts executive resource close to its ceiling. This is your reference point — not a population average, but your own best day.
The prefrontal cortex and hippocampus give way first — they are the heaviest consumers. The occipital holds. Look at the map: the regions that fall are not the ones you would guess.
No consumer competitor exposes its model. Here you click, and you see the arithmetic. The numbers add up — and you can check.
Recovery is asymptotic: most of it happens on the first night, the rest comes back slowly. The model says so, and states how far off it may be.
Most cognitive performance tools hand you a score and ask you to believe it. This one hands you the score, the formula, and the margin of error.
Sleep, stress, fatigue, inflammation, hydration, alcohol, nicotine, current medication. Nine sliders, twenty seconds. Nothing is measured on your body: this is a simulation built from what you report.
Each input weighs differently depending on the brain region. The directions follow established mechanisms; the magnitudes are an engineering choice the product owns and documents, version by version.
The timed baseline measures your actual performance. The gap between what the model predicted and what you did is displayed, not buried. That is what makes the tool arguable — and therefore credible.
Not ten wellness tips. A presence in the app: it reads your sliders, mirrors them back, and picks one step for the next 2 hours — computed on your device, not in a cloud.
Two minutes in the morning. Kato mirrors your numbers with a tone, then one commitment — not an Instagram checklist. That’s what “sleep 8 hours” apps never do.
Simulated check-in · stress ~2/10
You said stress is low. While it holds: lock 45 min before the next spike — one move, not a list.
Demo here. In the app, Kato reads your sliders — not this script.
This block is not a screenshot. These are the application's own formulas, running here: move the sliders and watch the arithmetic redo itself. Add up the points deducted and you get the score back. It is the only argument we have, and it can be checked without taking our word for it.
100− 18,4− 7,2− 2.1= 72.3
This total exists only today. The 7-day trial compares it to your next days — that is what you pay for.
Move the sliders: the arithmetic redoes itself in front of you. These are the application's formulas, not a mock-up. Indicative values, not measurements.
Because an exposed model is an attackable model. The moment you publish your weightings, you invite contradiction — and you have to be able to withstand it.
We prefer that risk to the alternative. An opaque score teaches you nothing: it tells you how you are without telling you what it comes from. And it is the “what it comes from” that lets you decide, tonight, whether that last hour of work is worth the hour of sleep it costs.
The model's version log is available inside the application, with what changed and why.
They measure real performance, on you, under the conditions of the moment. They are compared to no sample: they exist to establish your own baseline, and then to place your days against it.
Raw processing speed, and above all its variability — often more informative than the average.
Working memory: hold a sequence and give it back in reverse. The longest span you reach is your marker.
Visuospatial working memory, dependent on the parietal cortex and the hippocampus.
Respond to one signal, hold back on the other. The prefrontal brake, measured directly.
Holding the instruction over time. It is the first function to give way after a short night.
Sleep, stress, fatigue, declared medication: these are health data under European regulation. That is precisely why they never leave your device.
Everything is computed and stored in your browser. No server receives what you declare, your cognitive measurements or your history.
This home page counts its visits and the clicks on its buttons, anonymously and without cookies, to know where its readers come from and what interests them. The application itself measures nothing at all. Your account exists only to check whether your subscription is active.
Immediate and unconditional. CSV and JSON export available at any time: your data follows you if you leave.
The demo on this page stays open and complete, forever: the film, the sliders, the live math. What you pay for is the app — your baseline, your history, your trends.
All features unlock once you add a card (no charge for 7 days). At day 7: subscribe or cancel in one click — nothing is charged if you cancel first.
The details, for those who want to read them. Your account exists only to verify that your access is active — your health data never leaves the device, including after payment. One-click cancellation from your account, with no justification and no email to send. A 14-day right of withdrawal applies under European distance-selling law; we go further with a refund on request within 30 days. Prices are in euros, inclusive of tax for consumers.
In a market saturated with promises, a product that draws its own boundaries tells you two things: what it actually does, and that it does not need to sell you a dream to stand up.
Seven days with a card (no charge during the trial). The full app opens after you add a payment method.