Keyboard Polling Rate Test
Measure your key repeat rate and the delay before the first repeat, and time the gaps between keyboard events. A browser cannot read the USB polling rate.
This is your operating system's key repeat setting, not the keyboard's USB polling rate. A 125 Hz office keyboard and an 8000 Hz gaming one will show the same numbers here.
What this test measures
Polling rate versus what you see here
USB polling rate is how often the computer asks the keyboard for news. A poll 125 times a second is an 8 ms gap, 1000 times a second is 1 ms, and 8000 times a second is 0.125 ms. This page cannot see that clock. Key repeat rate is the operating system auto-repeat, in repeats per second. Event timing is the gap between KeyboardEvent timestamps. Neither number is a USB polling rate.
Does 8000 Hz matter?
Going from 125 to 1000 Hz cuts up to 7 ms of worst-case delay, which a fast player can feel. Going from 1000 to 8000 Hz saves under 1 ms and costs CPU time, since the PC handles up to eight times more USB interrupts. The box number also says nothing about the scan rate (how often the keyboard checks its own switches) or debounce time. A board that polls at 8000 Hz but scans at 1000 Hz with 5 ms of debounce is not faster than a good 1000 Hz board. Look for reviews that measure end-to-end latency instead.
Changing the repeat rate and the polling rate
Windows: Control Panel, Keyboard, then the Repeat delay and Repeat rate sliders (about 250 to 1000 ms of delay and roughly 2.5 to 30 repeats per second). macOS: System Settings, Keyboard, then Key repeat rate and Delay until repeat; the fastest setting gives roughly 30 to 33 Hz. The polling rate is set in the maker's software, such as Razer Synapse, Logitech G HUB, Corsair iCUE or Wooting, or sometimes with a key combination in the manual. Many wireless modes drop it to save battery.
Related tools
Check every key works with the keyboard tester, see how many keys register at once with the keyboard ghosting test, catch switches that chatter with the keyboard double click test, and check your mouse with the mouse polling rate test.
Frequently asked questions
What does this keyboard test measure?
Two different things. The repeat test measures auto-repeat from the operating system: the delay before the first repeat, then how many repeat events arrive per second. Event timing records timestamps of distinct key presses and releases during a short roll. Neither one reads the USB polling rate.
Can a browser read the USB polling rate?
No. The page only receives KeyboardEvents. It cannot ask the USB host how often it polls the keyboard, so it does not report a USB polling rate. Hertz on the repeat test means repeat events per second, not the polling clock.
How is the repeat rate in hertz calculated?
The live number is 1000 divided by the median gap, in milliseconds, across recent auto-repeat events. The average uses the gaps kept for that hold. A gap of 0 ms is skipped, so the readout stays blank instead of becoming infinite. The repeat count is how many auto-repeat events arrived, not how many times you pressed the key.
Which events count as repeats, and which count as presses?
On the repeat test, the first press only starts the clock. Later events that the browser marks as repeats are the ones timed and counted. On the event timing test, auto-repeat events are ignored. The press total counts distinct key-down events. Key-up events are extra timing samples: they are included in the key-event total, not in the press total.
Why does Space not scroll, and how do I stop?
While the repeat test is listening, Space and the arrow keys are kept from scrolling so a held key can be measured. Release the key to stop, or press Reset to clear the numbers and ignore that key until the next press. Enter still activates buttons and links. On the event timing test, press Stop or Escape. Leaving the page removes the listeners.
Why might the repeat test show no repeats?
Key repeat may be turned off in the operating system. On macOS, holding a letter can open the accent menu instead of repeating. Hold Space or an arrow key for about two seconds. If event timing has too few distinct gaps, or the browser rounds every timestamp to the same value, that test says it cannot tell.
Will a faster keyboard change the repeat number?
Usually no. The repeat rate follows the operating system setting, often near 30 repeats per second at the fastest slider. A keyboard that polls faster can show the same repeat rate as a basic one. Use the repeat test to check that setting. Use event timing only as a look at gaps between events.