Keyboard polling rate test

How to test your keyboard polling rate

1. Press Start 10-second test. The 10-second countdown begins with your first key press.
2. Roll your fingers across as many different keys as you can, as fast as you can. Think of drumming on the whole keyboard rather than typing words.
3. After 10 seconds you see the result, for example ≈ 125 Hz with "8 ms between reports", how many intervals were measured and what share of them fit the pattern.
4. Press Run again to repeat. Two or three runs that agree are more convincing than one.

The chart under the result shows the gaps between your key presses in milliseconds. On a 125 Hz keyboard you can often see the gaps bunch up at 8, 16, 24 ms and so on.

How the estimate works

A keyboard does not tell the computer about a key at the instant you press it. It reports at a fixed rhythm, the polling rate: 125 Hz means a report every 8 ms, 500 Hz every 2 ms, 1000 Hz every 1 ms. Two presses that land in different reports are therefore separated by a whole number of report periods.

The test records the time between presses of different keys (repeats from holding a key are ignored). It then checks the periods 16, 8, 4, 2 and 1 ms and picks the largest one that at least 90% of your gaps fit, allowing ±0.25 ms of wobble. If 90% of the gaps are close to multiples of 8 ms, the keyboard almost certainly reports every 8 ms: ≈ 125 Hz.

• ≈ 125 Hz (8 ms): typical for office keyboards and many laptops.
• ≈ 250 Hz (4 ms) or ≈ 500 Hz (2 ms): some gaming and wireless keyboards.
• ≥ 1000 Hz (browser limit): 1 ms is the finest step this method can check, so 1000 Hz, 4000 Hz and 8000 Hz keyboards all land here.
• No clear pattern: the gaps do not line up with any fixed period. The system or the keyboard (common with Bluetooth) adds jitter. Try again with fast rolls over many keys.

Why a browser can only estimate it

A web page only sees when each key event arrived, after the USB stack, the operating system and the browser have handled it. Each of them adds a little delay and jitter. Some browsers also round event times on purpose to protect privacy (Firefox and Safari can do this, especially with anti-fingerprinting settings on). When the test detects such a coarse clock, it says so and shows no number, instead of inventing one.

That is why the result is labelled Estimate. It is good at telling a 125 Hz keyboard from a 1000 Hz one, and it shows you the data it used. It cannot measure the keyboard's own hardware latency, and it cannot tell 1000 Hz from 8000 Hz. For exact figures you need a USB analyser or the maker's own software. Tools that print a precise number like "997.3 Hz" from a browser are guessing.

Many gaming keyboards let you set the polling rate in their software or with a key combination. If you changed it, this test is a quick way to check the setting took effect.

Polling rate vs keyboard latency

People often search for a "keyboard latency test", but polling rate is only one part of latency. Going from 125 Hz to 1000 Hz cuts the worst-case wait for the next report from 8 ms to 1 ms. Switch debounce, the keyboard's scanning, wireless links, the operating system and your screen add their own delays, and a browser cannot separate them. If you want a feel for your total reaction chain, the reaction time test measures you plus your whole setup.

Related tests: the mouse polling rate test does the same job for your mouse with a different method, the keyboard test checks every key, and the rollover test shows how many keys register at once. Our guide to reading test results explains Hz and ms in more detail, and the laptop keyboard guide helps with keys that don't respond. Nothing you press is stored or sent. More tools: device tests.