KPS test
Hit any keys as fast as you can and see how many you land each second. You get your average and peak keys per second, a live graph, how your presses split between keys, and how evenly they are spaced. It suits rhythm game players and anyone tracking their own tapping speed.
Switch sounds are simulated from the switch specs, not recorded.
Keys per second
0
presses in the last second
- total presses
- 0
- time left
- 10.0 s
- last key
- -
Press any key, or tap the two keys below. The clock starts on your first press.
No best saved for 10 seconds yet.
What the test measures
KPS is keys per second: presses divided by the time they took. The large number during a run is a rolling count of the presses in the most recent second, so it rises and falls as you tap. Your average is the total divided by the test length, and your peak is the most presses inside any one-second window. Every key counts except Tab, Shift, Control, Alt, Meta and Caps Lock, and a key that is held down counts only once.
Pick 5, 10 or 30 seconds, or use the live meter, which runs until you press Esc or Finish. The timer starts on your first press and the time left is calculated from that moment, so it does not drift.
Reading the result
Stream BPM expresses your average as the tempo of a stream of quarter-beat notes, which is how rhythm game players usually quote tapping speed. There are four notes to a beat, so BPM = KPS x 15, and 12 KPS is a 180 BPM stream.
The consistency figure is the standard deviation of the gaps between your presses, in milliseconds. A lower value means more evenly spaced presses. It is not osu! unstable rate, which measures how far your hits land from the beat of a map. This page has no beat to compare against, so it cannot calculate that.
The key table counts presses per key. If you alternate two keys, the two shares show whether one finger is doing more of the work, and the alternation figure tells you how often you switched keys.
What limits your KPS
Stamina and technique come first, then hardware. A light linear switch with a short stroke is easier to repeat than a heavy tactile one, and the best switches for gaming page compares the usual choices. Firmware debounce and USB polling rate set how quickly a press is reported, covered in polling rate explained, and Hall effect switches with rapid trigger can reset a key sooner than a mechanical contact.
A board with limited rollover can drop presses when several keys are down together, which the ghosting test will show. If your keyboard resolves conflicting keys such as left and right pressed at once, the SOCD test shows what it does with them.
Limits of this test
This counts key events delivered to a web page, not what a game accepts. Games read input in their own way and apply their own timing rules, so your result here is a guide to your tapping speed, not a prediction of your score. Taps on the two on-screen keys measure how fast you can tap a screen, and are saved as a separate best. Different keyboards debounce differently, so compare runs on the same board. The switch you choose plays a simulated sound and has no effect on the count.
To count one key only, use the spacebar counter. If a key sometimes types twice, run the chatter test. The latency test measures the part of input delay a browser can see, and the typing test measures words per minute. If a key acts up, the key fix guides may help, and the keyboard tester checks that every key registers.
Frequently asked questions
What is a good KPS?
There is no single good number. It depends on the game, your key layout and how long you need to hold the speed. The useful comparison is your own earlier runs at the same length, which is why the test keeps a best for each one.
Is KPS the same as CPS?
The arithmetic is identical: presses divided by seconds. CPS usually means mouse clicks and KPS means key presses. This page counts keyboard keys, so it is not a mouse click test.
Why multiply KPS by 15 to get BPM?
A stream is a run of quarter-beat notes, four to every beat. Notes per second work out to BPM times 4 divided by 60, which simplifies to BPM divided by 15, so KPS times 15 gives the tempo of a stream at that speed.
Why is my 30 second score lower than my 5 second score?
A short burst and a long run test different things. Few hands can hold their top speed for half a minute, so a lower average over 30 seconds is expected. Compare like with like.
Why are Shift and Tab not counted?
They, along with Control, Alt, Meta and Caps Lock, are used to operate the page and to move around it by keyboard. Every other key counts, and a key held down counts once.