Reaction box
Space to arm, Space on amber
--
Click here, then press Space to arm the test
Click the box, press Space to arm it, wait for the colour change, then press Space as fast as you can. Five rounds give an average that combines your reflexes with the delay of your keyboard, USB or wireless link and operating system. Run the keyboard test to see your result in seconds, right in your browser.
Space to arm, Space on amber
--
Click here, then press Space to arm the test
Use this keyboard latency test to see how many milliseconds pass between your key press and the moment your browser registers it. Press any key and the tool logs your keyboard latency, your input lag and an estimated scan rate, so you can tell a sluggish board from a sharp one in seconds. The readout also tracks your response time across presses, which matters most for gaming and fast typing.
This online keyboard latency test runs entirely in your browser, so there is nothing to install. It records live telemetry for every key press, including processing lag and event timing, and plots the results as a histogram. Any keyboard latency checker works on the same principle: you generate key events and read how long each one took.
Run this keyboard delay test in a few quick moves. The Key Latency view shows how long the browser takes to process each press, while the Event Timing view graphs the gap between consecutive key events. Lower, steadier values mean less input lag. The test is free and needs no signup. Plug the keyboard straight into a USB port on the computer, because a USB hub can add its own wait.
Every press adds a row to the key press history. The shortest key press in that list gives your estimated keyboard scan rate, and the same list yields your minimum latency, maximum latency and average latency. Trust the average, because a single spike usually comes from system activity rather than the board.
$$\text{Scan rate (Hz)} = \frac{1000}{\text{shortest key press (ms)}}$$
A shortest press of 3 ms works out to 1000 ÷ 3, or roughly 333 Hz.
A keyboard latency tester is most useful when something changed or something feels off:
Pair the result with a separate reaction time test if you want to see how much of your delay is the board and how much is you.
Keyboard latency is the total delay between physically pressing a key and your computer acting on it. Lab measurements run end-to-end, from the key starting to move to the screen updating (keypress-to-screen), using a logic analyzer or a high-speed camera. A browser tool samples only the software side of that chain. Keyboard input latency is stated in milliseconds, and system latency from the operating system adds to whatever the hardware contributes. The mac keyboard test runs entirely in your browser, so nothing you press is stored or sent anywhere.
People use keyboard response time and latency as if they were the same thing. Response time describes the complete trip from press to visible output, so it includes the scan, the report to the PC and the on-screen update. Key response time is simply that figure for one press, and a higher response time always means a more sluggish feel.
Total keyboard latency is the sum of four delays:
$$\text{Latency} = t_{\text{debounce}} + t_{\text{scan}} + t_{\text{polling}} + t_{\text{system}}$$
The average wait for the polling interval is half the interval, so a keyboard reporting at 500 Hz (a 2 ms interval) adds 1 ms on average. Say a board needs a 4 ms debounce, a 1 ms scan, that 1 ms polling wait and 7 ms of operating system and browser processing: 4 + 1 + 1 + 7 = 13 ms. That lands in the typical band in the benchmark table below.
Four things set most of your keyboard latency, and together they shape keyboard performance: how often the board reports to the PC, how often it scans its keys, how it connects, and how busy your computer is. Each one has a different fix, so it helps to separate them. Run the rapid trigger test to see your result in seconds, right in your browser.
The keyboard polling rate is how often the keyboard reports its state to the computer, also called the report rate. Over USB, a 125Hz board reports every 8 ms, a 1000Hz board every 1 ms, and some gaming boards reach 8000Hz. The polling rate is measured in hertz while latency is measured in milliseconds, so the two are related but not interchangeable. A higher polling rate improves performance only by trimming the waiting time; going from 125Hz to 500Hz cuts the average wait from 4 ms to 1 ms.
The scan rate is how many times per second the firmware checks every key. After a switch actuation, the board waits out the debounce time so one press isn't counted twice, and that pause is baked into the latency. A mechanical keyboard often has a short actuation distance, while a membrane keyboard usually needs a deeper press. Features such as key rollover (NKRO) and anti-ghosting share the same matrix scan, so they matter when you hold several keys at once.
Your connection type changes both the average latency and how much it varies. A wired USB connection is the steadiest, while a wireless keyboard trades some consistency for freedom of movement.
| Connection | Typical latency | Note |
|---|---|---|
| Wired USB 2.0 | Low and stable | Fine for most gaming |
| Wired USB 3.0 | Low and stable | Use a direct port, not a hub |
| 2.4GHz wireless dongle | Close to wired | The wireless dongle sits in a USB port |
| Bluetooth | Higher and more variable | Best for office work and travel |
Software around the keyboard can add delay even when the hardware is fast:
I just moved to a 75% board that advertises a 1000Hz polling rate, and my first-person shooter flicks feel a touch late. The question is simple: is the delay the board or my setup, and does it clear the 8 ms I want for shooters? I plug it straight into a rear USB port, close the browser tabs, and tap twenty mixed keys.
The history shows a shortest key press of 2.5 ms, so the scan rate is 1000 ÷ 2.5 = 400 Hz. The average over the twenty presses comes out at 9.6 ms, which is typical but not strong. Before blaming anything, I rebuild the number by hand: a 3 ms debounce, half of the 2.5 ms scan interval (1.25 ms), 0.5 ms of average polling wait at 1000Hz and roughly 5 ms of browser and system time gives 9.75 ms. That is within 0.2 ms of the measured figure, so I trust the test.
What stands out is the debounce setting, not the polling rate. The vendor software lets me cut it, so I rerun the sums with one input changed at a time:
| Change | Estimated average | Verdict |
|---|---|---|
| As shipped | 9.75 ms | Typical band |
| Debounce 3 ms to 1 ms | 7.75 ms | Under my 8 ms goal |
| Polling dropped to 125Hz | 13.25 ms | Wrong direction |
The debounce cut wins by 2 ms, while the polling rate is already past the point of diminishing returns. I make that change, rerun the test and get an average of 7.9 ms, just inside the band. The edge case I watch for: a lower debounce can double-register a worn switch, so I type a paragraph afterwards to make sure no letters repeat. If they do, I raise it to 2 ms rather than trust the lower number.
One caveat keeps me honest. A single 31 ms spike appeared while a background update ran, so I judge the board on the average rather than the worst press, and the browser can't see the keyboard's internal scan time anyway. If the shooter still feels late, I would test the monitor next, because this reading covers only the keyboard side of the chain.
What counts as good depends on what you do. Gamers chasing a competitive edge want the lowest numbers, while office users rarely notice anything under about 10 ms.
Use these latency benchmarks to rate your average from the test. One 60 Hz display frame lasts about 16.7 ms, which makes a convenient yardstick.
| Average latency | Rating | Good for |
|---|---|---|
| Under 3 ms | Elite | Esports and rhythm games |
| 3 to 8 ms | Strong | Any gaming keyboard use |
| 8 to 16 ms | Typical | Everyday gaming and typing |
| Over 16 ms | Laggy | Check drivers and connection |
Not automatically. Independent measurements of keyboards show tens of milliseconds of spread between models, and gaming-branded boards weren't consistently the quickest. Marketing claims about polling rate rarely mention that debounce and firmware often dominate the total, so judge a board by measured latency rather than the box. Even a high USB polling rate can be undone by a slow connection.
A browser measures what the web page sees, which is the browser's event processing rather than the full hardware chain. Results shift with the browser, system load and open tabs, and the test can't see the keyboard's internal scanning. Treat the output as a dependable way to compare the same keyboard's latency before and after a change, not as a lab-grade benchmark. For exact accuracy, a dedicated hardware tool is the right instrument, and the responsiveness you feel in games is the final judge of performance.
For everyday typing, a few milliseconds are invisible, even at a high typing speed. People do notice small delays on simple tasks, though, and errors creep in as delay grows. Run a keyboard input lag test if letters seem to trail your hands, then compare the result with the table above.
If your average sits above where you want it, work through these fixes in order and re-run the test after each one to see which actually helped:
Fixing the cause of keyboard lag is usually a matter of one setting or one cable, so measure before you spend money on a new board.
Open a keyboard latency test and press any key. Each press is timed and listed in a key press history, and the shortest press gives an estimated scan rate. Repeat with several keys and compare the average.
Under 5 ms is excellent for competitive play, 5 to 10 ms is acceptable for most users, and anything above about 10 to 20 ms can feel laggy in fast games. Latency above 20 ms may affect rhythm games and fast typing.
Scan rate is how often the keyboard checks its keys, measured in hertz. Latency is the total delay from the press to the computer registering it, measured in milliseconds.
Add the debounce time, the polling interval (1000 divided by the polling rate in Hz) and the system latency. The calculator uses that sum as the maximum latency, and estimates scan rate as 1000 divided by the shortest key press time.
Not fully. A browser test measures event processing inside the browser, so results vary with system load and cannot include the keyboard's own internal processing. Dedicated hardware tools give exact totals.
Common causes are a Bluetooth or other wireless connection, outdated drivers or firmware, high CPU load and background processes, USB hubs, and Windows settings such as Filter Keys, Sticky Keys and power saving.
Use a wired or 2.4GHz connection plugged directly into a USB port, raise the polling rate in the keyboard software, update firmware and drivers, and close unneeded background applications.
No. 100 ms is very high and makes input feel noticeably delayed, especially in fast-paced games. Aim for 10 ms or less, with 5 ms or lower being the most responsive range.