Keyboard Rollover Test

How to use

1. Use Free mode to explore how many simultaneous key events reach the browser, or Guided combination mode when you want to test a known combination.
2. In Guided mode, choose a preset or build an expected combination on the virtual keyboard, then press Start guided round.
3. Press and hold all expected keys at the same time. Keep them held together briefly, then release all keys to complete the round.
4. Review whether the expected combination matched, plus Registered together, Possible missing together, Unexpected codes observed, and total versus non-modifier counts.
5. Repeat several combinations that matter to your actual use. Reset clears the current session's free peak, guided rounds, and custom combination.

What the result means

• Currently pressed shows the codes the browser presently considers held. The total count includes modifiers; the non-modifier count excludes Shift, Control, Alt, and Meta while treating Caps Lock as a regular key.
• The maximum simultaneous keys value is a browser-observed count, not a hardware rating.
• Guided best match is the largest number of expected codes that were present together at one instant. Registered together lists the expected codes in that actual state.
• Possible missing together lists expected codes not present in the best simultaneous state. Unexpected codes are browser codes outside the expected set that appeared during the controlled round.
• A successful round validates only that tested combination at the browser level. One successful six-key combination does not certify universal 6KRO or NKRO.
• Six-key rollover is commonly discussed alongside the USB keyboard boot-protocol limit, but this page does not identify the active report protocol or turn one six-key sample into a device-wide rating.
• Rollover describes keys reported together; N-key rollover describes a broader ability to report arbitrary combinations. Blocking means an expected key is absent, while ghosting traditionally describes an extra phantom key. This page reports possible missing and unexpected browser codes instead of assigning either hardware cause automatically.
• A missing code alone cannot prove whether the cause is the keyboard matrix, firmware, connection/reporting protocol, OS shortcut, browser/software handling, focus, or interruption.
• The guided analysis listens to browser keydown/keyup events, uses KeyboardEvent.code, ignores repeat keydowns for concurrency, replays the round chronologically, and reconstructs which codes were down together; it does not union every code seen during the round.
• Rollover describes simultaneous key-reporting behavior. NKRO is a broad capability claim, not a score this page can certify; ghosting or blocking language describes observed combination behavior, not a proven matrix fault.
• This is browser-observed rollover evidence, not electrical NKRO certification; this browser result is not NKRO certification, and a possible missing browser event is not automatically a hardware fault.

Limitations

• Browser results show only key events delivered to this page. A successful combination does not certify all combinations or full NKRO, and a missing event does not by itself prove a hardware fault.
• The operating system or browser may intercept shortcuts before their key events reach the page.
• Choose custom expected keys on the on-page keyboard before pressing them physically. This preserves an intended key as expected even when its event never reaches the browser.
• The page cannot independently verify which physical keys your fingers touched, so an unexpected browser code may be accidental, remapped, software-generated, or phantom-like without proving hardware ghosting.
• Different combinations can behave differently. Browser-observed evidence is not a full hardware matrix, firmware, report-protocol, or electrical certification.

Troubleshooting

A shortcut combination is incomplete

Try a combination that is not reserved by the operating system or browser, keep this tab focused, press all expected keys together, release all keys, and repeat the same guided round.

The result says Possible missing even though every key worked individually

Individual registration and simultaneous rollover are different tasks. Repeat the exact combination; a keyboard can report each key alone yet fail a particular multi-key combination.

Guided results change between runs

Press all expected keys together rather than sequentially, hold them briefly, then release all. Compare repeated identical combinations under the same focus and browser conditions.

A round was interrupted

Focus loss or page hiding invalidates the guided round. Return to the page and start a fresh round; interrupted rounds are not classified.

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 Event Timing Test

  Compare browser-received keyboard event timing with controlled tapping, alternating-key, and OS auto-repeat runs. Review intervals, cadence, consistency, and local history—not USB polling.

• 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