Click Speed Test to the millisecond

It counts pointer presses and divides by the elapsed time. That sounds trivial, and the details are where most click speed tests go wrong: the clock has to start on your first press rather than on page load, or reaction time is charged against your rate; the interval has to be closed by a timer rather than an animation frame, or a backgrounded tab runs long; and elapsed time has to be clamped to the nominal interval, or the projected rate drifts.

All three are handled here, which is why two runs of the same length are comparable. The clock source and its resolution are printed under the instrument so you can see what is being claimed.

Short intervals measure your peak and long ones measure what you can hold. One second is a pure burst, five is the window a tense technique survives, ten is the standard, thirty exposes fatigue, and sixty is the only interval where clicks per minute is a count rather than a projection.

A rate that holds across two different intervals is a rate you actually have. If your ten-second figure is far above your thirty-second one, the higher number is a peak, and quoting it as your click speed is quoting your best moment rather than your ability.

In order: technique, then conditioning, then hardware. A single finger reaches 180 to 400 CPM, jitter 450 to 700, and a clean butterfly 600 to 950 — differences that dwarf anything a mouse contributes. Conditioning decides how much of that survives past thirty seconds.

Hardware matters only at the top. A switch's reset time decides how soon the button can register again, and its debounce window puts a floor under the interval between two accepted presses. None of it matters at 300 CPM; all of it matters at 900.