Explainer

Keyboard Latency: Where Input Lag Really Comes From

Between your finger starting to press a key and a character appearing on screen lies a pipeline of a dozen delays — mechanical travel, debounce, matrix scanning, USB polling, operating system queues, application processing, and display refresh. Marketing focuses on one or two of these numbers. This guide walks the whole pipeline, with realistic magnitudes, so you know what is worth optimizing and what is noise.

KeyboardAnalyzer TeamLast updated: August 24, 2026

The input pipeline, stage by stage

A keypress travels through these stages, each adding delay: physical key travel to the actuation point (highly variable — this is where switch choice matters), switch debounce (the firmware waits a few milliseconds to confirm a clean contact, typically 1–10 ms), matrix scanning (the controller sweeps the key grid, adding roughly 0.1–1 ms), the connection hop (USB polling or a wireless link), and finally the operating system and application, which queue and process the event before anything reaches the screen.

End to end, a good wired gaming keyboard contributes somewhere between 2 and 15 ms from actuation to USB report. For context, a 60 Hz monitor adds up to 16.7 ms per frame just in refresh, and game rendering pipelines typically add several frames more. The keyboard is usually a minority share of total input lag — which is exactly why measuring, rather than assuming, is worthwhile.

Polling rate: what 1000 Hz really buys

USB keyboards do not push events whenever they like — the host polls them on a schedule. At 125 Hz, the poll interval is 8 ms, meaning a keypress waits up to 8 ms (average 4 ms) for the next poll. At 1000 Hz the worst case drops to 1 ms. Moving from 125 Hz to 1000 Hz therefore saves an average of about 3.5 ms — real, but small.

Recent gaming keyboards advertise 4000 or 8000 Hz polling. The additional saving over 1000 Hz is under a millisecond — far below human perception thresholds — while costing measurable CPU overhead. Sub-millisecond polling is a spec-sheet race, not a felt improvement. 1000 Hz is the sensible target; beyond it, spend your money elsewhere in the pipeline.

Switches, actuation and debounce

Switch choice affects latency mainly through actuation distance: a switch that actuates at 1.0–1.2 mm (speed switches) fires measurably earlier in the press than one actuating at 2.0 mm — at typical press speeds, roughly 5–10 ms earlier. This is one of the larger levers in the keyboard itself, though it comes with more accidental presses.

Analog hall-effect and optical switches take this further with adjustable actuation points and rapid-trigger behavior (re-arming the key the instant it starts moving up), which meaningfully tightens repeated presses in games. Optical switches also skip traditional debounce, saving a few more milliseconds; conventional switches depend on the firmware's debounce algorithm, which ranges from about 1 ms in good implementations to 10+ ms in sloppy ones — a bigger spread than most switch differences.

For typing rather than gaming, none of this is perceptible. Latency optimization is a gaming concern; typists should choose switches for comfort and accuracy.

Wired vs. wireless: the honest comparison

Modern 2.4 GHz dongles from major gaming brands achieve latency statistically indistinguishable from their wired mode — the good ones poll at 1000 Hz over the air. A quality wireless gaming keyboard is no longer a compromise.

Bluetooth is a different story: connection intervals commonly add 10–20 ms and can spike under interference or when the radio power-saves. Bluetooth is fine for productivity and unsuitable for competitive play. If your keyboard offers both modes, this is one of the few settings changes that produces a genuinely measurable improvement: use the dongle or cable for gaming.

Beware of battery-saving modes on wireless boards — some raise effective latency when idle to stretch battery life, making the first keypress after a pause noticeably slower.

Measuring what you actually have

A browser-based keyboard performance test measures the intervals it can see: how quickly repeated presses register, the consistency of event timing, and effective event throughput. It cannot see the mechanical stage (finger contact to actuation) — full-pipeline measurement needs a high-speed camera or dedicated hardware — but the browser-visible portion is precisely the part where polling problems, wireless spikes, and OS interference show up.

What to look for in results: consistency matters more than the average. A keyboard that delivers events at a steady 2–4 ms is better for timing-sensitive play than one averaging 3 ms with 20 ms spikes. Spikes point to wireless interference, USB hub contention, or background software. Test in your real setup — same port, same connection mode, same running apps — and retest after changing one variable at a time.

What actually reduces your input lag

  • Use a wired connection or a 2.4 GHz dongle, never Bluetooth, for competitive play — saves 10–20 ms over Bluetooth.
  • Plug the keyboard directly into the PC, not through an unpowered hub or a monitor's USB passthrough.
  • Make sure the keyboard runs at 1000 Hz polling (check the manufacturer software); beyond 1000 Hz is placebo.
  • If your board supports it, reduce actuation distance or enable rapid trigger for games — one of the few genuinely felt keyboard-side changes.
  • Attack the bigger delays outside the keyboard: a high-refresh monitor, exclusive fullscreen, and capped-but-stable frame rates each save more milliseconds than any keyboard upgrade.
  • Close background software that hooks keyboard input (some RGB suites, macro tools, and overlays add measurable delay).

Measure your keyboard's response

Run the free performance test to see your keyboard's real event timing and consistency — then change one variable and test again.

Start the Performance Test

Frequently Asked Questions

Can humans even feel a few milliseconds of input lag?

Not as isolated events — blind tests show most people cannot distinguish single-digit millisecond differences. But delays add up across the pipeline, and consistency is perceptible: irregular timing feels 'floaty' even when the average is low. Cutting 10–20 ms of total system lag is noticeable; chasing 0.5 ms on one component is not.

Do optical switches make a keyboard faster?

Slightly and genuinely: they eliminate debounce delay (a few milliseconds) and often actuate on shorter travel. Whether that is felt in play is doubtful, but combined with rapid-trigger features they do tighten fast repeated inputs. Their bigger practical advantage is switch longevity — no metal contacts to wear and chatter.

Why do my browser test results vary between runs?

The browser shares the machine: background tabs, compositor work, and CPU load all jitter event timing. For comparable results, close other tabs and apps, use the same browser, repeat the test a few times, and compare medians. A consistent difference across repeated runs — for example Bluetooth vs. dongle — is real; run-to-run wobble of a millisecond or two is measurement noise.

Is a 'gaming keyboard' actually faster than an office keyboard?

Usually yes, but modestly: gaming boards typically poll at 1000 Hz (vs. 125 Hz on office boards), use faster debounce, and offer shorter actuation. Total difference: perhaps 5–15 ms. Worth it for competitive play, irrelevant for e-mail — and a good office keyboard with clean, consistent key registration is a better typing tool than a poor gaming one.

Related guides

All guides