Keyboard Event Timing Test

How to use

1. Choose Single-key tap cadence to repeatedly tap one ordinary key, Alternating F / J cadence to alternate F and J, or OS auto-repeat to hold one key and observe operating-system/browser repeat events. Select 5, 10, or 15 seconds.
2. Focus the capture surface and press Start. Keep the tab focused for the whole run so the browser can deliver the events consistently.
3. Tap an ordinary key such as A in single-key mode, alternate F and J in pair mode, or hold one key in repeat mode.
4. For tap or alternating-key comparisons, collect at least 20 accepted intervals. For repeat comparisons, collect at least 10 steady repeat intervals. These are recommended browser-event samples, not automatic stopping rules.
5. Let the timer finish or press Stop to freeze the current result before inspecting it. Reset clears only the current capture and target.
6. For comparisons, repeat the same mode and duration with the same keyboard, connection, browser, operating-system settings, and system load.

What the result means

• Single-key and alternating-key results describe human tapping cadence as delivered through browser keydown events. The primary rate is 1000 / median accepted interval: median is the representative middle gap, mean is the average gap, P95 is the slower tail, and minimum is the fastest single accepted gap—not a complete characterization by itself.
• Coefficient of variation is relative interval variation within the run; a lower value means the accepted intervals are more consistent. Repeat mode reports an OS/browser auto-repeat stream rather than USB polling.
• Single-key mode follows the first non-repeat keydown. Pair mode accepts a strict alternating F↔J sequence. Repeat mode isolates one target key's repeat events and reports its repeat interval, initial delay, and interval variation.
• These are browser-observed KeyboardEvent timings. They include delivery and scheduling behavior; repeat mode reflects OS/browser auto-repeat. They are not USB polling, HID report frequency, keyboard scan rate, debounce, actuation, or end-to-end latency measurements.
• Hardware polling reference — not measured here: theoretical report intervals are 125 Hz → 8 ms, 250 Hz → 4 ms, 500 Hz → 2 ms, 1000 Hz → 1 ms, 2000 Hz → 0.5 ms, 4000 Hz → 0.25 ms, and 8000 Hz → 0.125 ms. These are not expected browser values, a score, or proof of hardware behavior.

Limitations

• The measurement starts with KeyboardEvent records after keyboard firmware, the connection, the operating system, and the browser have already processed input.
• Human tapping cadence is part of single-key and alternating-key results; these modes are not hardware polling measurements.
• Auto-repeat mode is generated by operating-system and browser repeat timing, including its initial delay and steady interval.
• Window focus, browser scheduling, system load, operating-system settings, and connection conditions can change which events arrive and when.
• The page cannot observe raw USB/HID reports, matrix scan rate, debounce time, switch actuation latency, controller firmware timing, or end-to-end input latency.

Troubleshooting

The rate is blank

Focus the capture surface and produce enough accepted events for the selected mode. Reset and try again without switching tabs.

Browser cadence does not match an advertised keyboard rate

This browser test is not expected to reproduce a keyboard's advertised hardware polling rate. Tap modes include human cadence, and repeat mode reflects operating-system/browser auto-repeat.

Repeat mode differs between computers

Repeat mode observes operating-system and browser auto-repeat behavior. Match the OS repeat settings, browser, and focus conditions before comparing runs.

Results vary between runs

Use the same mode, duration, key sequence, keyboard connection, browser, operating-system settings, and system load. Compare the aggregate local history entries after each stopped run.

Related tools

• Keyboard Tester

  Test keyboard keys online across common Windows, Mac, and compact layouts. See pressed and detected keys, event.key/event.code, repeat state, simultaneous browser input, and untested keys.

• Keyboard Rollover Test

  Test keyboard rollover with free and guided key combinations. See max simultaneous keys, best match, possible missing or unexpected codes, with clear NKRO and ghosting limits.

• Keyboard Chatter Test

  Test keyboard chatter, double typing, and unwanted double presses with controlled keydown/keyup timing. Review separate browser-level flags while auto-repeat is excluded.

FAQ