Measure the speed of you.

What these tests actually measure

KeenMind Lab is a small lab for a handful of human performances you can finish before a kettle boils. The headline number on each page is one thing: simple visual reaction time, clicks per second, aim latency, words per minute, a memory span, a spacebar rate, a hardware diagnostic, or a 2048 score. We do not mix them. A “fast” person at waiting-for-green is not automatically a fast typist, and a high CPS does not mean a healthy mouse. The point of a hub is that you can walk from one instrument to the next without being sold a personality type.

The speed tests — reaction time, click speed, aim trainer, typing, and the spacebar speed test — measure a loop from stimulus or intent to a switch. Reaction time is the cleanest of those: wait for green, click, average five valid trials. CPS and spacebar count how many times you can close a switch in a window. Aim adds a spatial target. Typing adds language. None of them is an IQ score. They are clocks.

The memory tests — sequence memory, number memory, visual memory, and the chimp test — measure how much you can hold for a couple of seconds and put back. Sequence is order. Number is a digit string. Visual is a spatial set. The chimp test is both: numerals in space, then a tap sequence after they hide. They feel related and they are not interchangeable. People who clear level 12 on sequence often stall at 8 digits on number memory. That gap is the interesting part.

Two pages are not games so much as desk diagnostics. The keyboard tester lights every physical key so you can find the switch that died. The double click test flags intervals under 80 milliseconds, the signature of a bouncing mouse switch. They bring in a different search intent — “is my hardware broken” — and then those people wander into the games. KeenMind Lab is happy about that. And 2048 is the puzzle anchor: a sliding-tile board in the same dark-violet language, with undo and a win at 2048. It is slower on purpose. Not everything here has to be a sprint.

Typical scores, side by side

These ranges are the middle of the distributions we use for percentiles — not a promise, not a medical chart. They exist so a first visit does not feel like an empty gym. Your personal best, once you have one, replaces the typical line on each card.

Reaction time in this table is simple visual RT on a browser, centered around 248 ms. Competitive players on high-refresh displays sit lower; tired people and laptop trackpads sit higher. CPS is the 5-second standard. Spacebar SPS runs a little cooler than mouse CPS because the switch is larger. Typing is net WPM on common English, not code. Memory “typical” is the level or span where a first-session adult usually lands. 2048’s “typical” is reaching the namesake tile, not a point total — point totals wander with leftover merges.

How the percentile system works

When a run ends you see a bar: “you’re faster than 73% of people.” That number is not scraped from this site’s traffic. A hub of CPS tests at 1 a.m. is a biased sample. Instead each game has a hardcoded curve — a mean and a spread — shaped like published simple-RT tables, large public typing corpora, and the online datasets that have accreted around these toys for a decade. Your score is placed on that curve. The bar is capped at 1% and 99% so nobody is told they are the fastest organism alive.

Direction matters. On reaction time and aim, lower is better, so a small number maps to a high percentile. On CPS, spacebar, typing, memory, and 2048, higher is better. Diagnostic tools (keyboard tester, double-click) still report a number because the rest of the site does, but the useful sentence is the verdict, not the percentile. Durations have their own curves: a 1-second CPS of 10 is ordinary; a 60-second CPS of 10 is not. That is why the 1-second and 60-second pages are separate URLs instead of a dropdown nobody can search.

Personal bests live in localStorage on this device. We do not upload them, we do not rank you against other visitors, and we do not ask for an email. If you want a comparison, send a friend the link. If you want the lab to keep the lights on, there is a discreet support link in the footer. Ads, when they are approved, sit on result screens and in the articles — never inside a play area, never next to a click pad. That is not a slogan. Accidental ad clicks are how sites like this get banned.

FAQ