Observed event ceiling
The fastest consistent received gaps are converted into an estimated events-per-second ceiling rather than relying on one accidental timestamp.
Run a quick online keyboard Hz check. See the observed browser-event ceiling, nearest common rate, shortest and median gaps, jitter, stability and recent key events.
Start the timer, then tap A and S alternately in short, quick bursts. The tool uses the fastest consistent gaps received by this page to estimate an event-rate ceiling.
Start, then tap A and S alternately as quickly and evenly as you can.
Keyboard timing stays in this browser.
No key event yet
A browser can estimate the rate of keyboard events that reach a web page, but it cannot directly read the keyboard's USB report descriptor or measure every hardware poll. Tapping speed, the operating system, browser scheduling and scan behaviour all affect the result. Treat the displayed Hz as a playful browser-event estimate and compare repeated runs on the same setup—not as proof of a keyboard's advertised polling rate.
The page keeps the quick interaction people expect while exposing the timing behind the estimate and the reasons it can move.
The fastest consistent received gaps are converted into an estimated events-per-second ceiling rather than relying on one accidental timestamp.
Compare the sample with familiar 125, 250, 500, 1000, 2000, 4000 and 8000 Hz labels and their theoretical intervals.
Shortest gap, median gap, jitter and a simple stability score make an uneven sample visible.
Inspect accepted keys and timestamp gaps. Held-key repeats and modifier shortcuts are left out of the sample.
A polling rate and the event rate visible to a page are related ideas, but they are not the same measurement. Read the whole sample before comparing two runs.
A hardware report rate is commonly expressed in hertz. A theoretical 1000 Hz interval is 1 ms; 500 Hz is 2 ms and 125 Hz is 8 ms.
One second divided by the average of the fastest consistent event gaps. It reflects what the page received during your tapping pattern.
Jitter is the spread of accepted event gaps. A steadier tapping rhythm and a quieter system can produce a more consistent sample.
The closest common label is shown only as a reference. It does not identify firmware, scan rate, connection mode or verified USB reports.
A web page receives keyboard events after the device, connection, operating system and browser have processed them. It cannot distinguish an 8000 Hz USB report stream from the keyboard's internal scan rate using ordinary keydown events. For verified hardware testing, use manufacturer diagnostics or dedicated USB measurement equipment. This page is for fun and repeatable same-device comparisons.
Public web standards document the event signal available to this page; the USB-IF reference shows why hardware report testing is a separate layer.
Reviewed 29 July 2026Answers about checking Hz, 1000 Hz and 8000 Hz, latency, changing results, wireless keyboards and what a browser can actually measure.
Start the 10-second test and tap A and S alternately in quick, even bursts. Repeat the run two or three times in the same browser and compare the observed ceiling and gap stability.
It commonly describes how often a keyboard can report input to the computer, measured in hertz. A 1000 Hz rate has a theoretical report interval of 1 millisecond.
1000 Hz is a common gaming rate with a 1 ms theoretical interval. Actual input latency also includes switch scanning, debounce, connection, operating-system processing, game timing and display latency.
It can change the waiting time between possible hardware reports, but it is only one part of total latency. A higher label does not guarantee a faster complete input path.
No. Ordinary browser keydown events cannot certify an 8000 Hz USB report stream. The page may show very short received gaps, but verification needs device-level diagnostics or dedicated hardware.
Your tapping rhythm, browser scheduling, background load, power mode and connection can all change the received gaps. Keep the setup fixed and compare several runs instead of one peak.
Yes, for a same-setup browser comparison. Keep the receiver position or Bluetooth connection unchanged. The page cannot separate wireless transmission from the rest of the input path.
Use only settings and firmware provided for your keyboard model. Check its official software, connection mode and documentation. A website cannot change hardware polling rate, and forcing unsupported settings may make input unstable.