ByteScope

Keyboard Tester

Press every key, watch it light up — and catch rollover limits and chattering switches.

Your files never leave your browser — all processing is local.

0 / 84 keys tested

Esc
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
F11
F12
`
1
2
3
4
5
6
7
8
9
0
-
=
Backspace
Tab
Q
W
E
R
T
Y
U
I
O
P
[
]
\
Caps Lock
A
S
D
F
G
H
J
K
L
;
'
Enter
Shift
Z
X
C
V
B
N
M
,
.
/
Shift
Ctrl
Meta
Alt
Space
Alt
Meta
Menu
Ctrl
Ins
Home
PgUp
Del
End
PgDn
↑
←
↓
→

Modifier state

CtrlShiftAltMetaCapsLockNumLock n/a

NumLock state isn't reported by every platform/browser — "n/a" means unavailable, not off.

N-key rollover

Currently held
0
Session max
0

6KRO keyboards guarantee 6 simultaneous non-modifier keys before drops; NKRO reports every key held at once. Hold down as many keys as you can — the session max is your hardware's real rollover.

Key chatter

No chatter detected yet — a flagged key's keyup→keydown gap was under 30ms.

Latest event

Press any key to see its details.

Event log

No events yet.

Full stress-test report

0 / 84 keys tested (0%) on ANSI TKL

Untested keys (84)

Escape, F1, F2, F3, F4, F5, F6, F7, F8, F9, F10, F11, F12, Backquote, Digit1, Digit2, Digit3, Digit4, Digit5, Digit6, Digit7, Digit8, Digit9, Digit0, Minus, Equal, Backspace, Tab, KeyQ, KeyW, KeyE, KeyR, KeyT, KeyY, KeyU, KeyI, KeyO, KeyP, BracketLeft, BracketRight, Backslash, CapsLock, KeyA, KeyS, KeyD, KeyF, KeyG, KeyH, KeyJ, KeyK, KeyL, Semicolon, Quote, Enter, ShiftLeft, KeyZ, KeyX, KeyC, KeyV, KeyB, KeyN, KeyM, Comma, Period, Slash, ShiftRight, ControlLeft, MetaLeft, AltLeft, Space, AltRight, MetaRight, ContextMenu, ControlRight, Insert, Home, PageUp, Delete, End, PageDown, ArrowUp, ArrowLeft, ArrowDown, ArrowRight

Keyboard stress-test report — bytescope.dev/keyboard-test

Layout: ANSI / TKL
Completion: 0/84 keys (0%)
Untested keys (84): Escape, F1, F2, F3, F4, F5, F6, F7, F8, F9, F10, F11, F12, Backquote, Digit1, Digit2, Digit3, Digit4, Digit5, Digit6, Digit7, Digit8, Digit9, Digit0, Minus, Equal, Backspace, Tab, KeyQ, KeyW, KeyE, KeyR, KeyT, KeyY, KeyU, KeyI, KeyO, KeyP, BracketLeft, BracketRight, Backslash, CapsLock, KeyA, KeyS, KeyD, KeyF, KeyG, KeyH, KeyJ, KeyK, KeyL, Semicolon, Quote, Enter, ShiftLeft, KeyZ, KeyX, KeyC, KeyV, KeyB, KeyN, KeyM, Comma, Period, Slash, ShiftRight, ControlLeft, MetaLeft, AltLeft, Space, AltRight, MetaRight, ContextMenu, ControlRight, Insert, Home, PageUp, Delete, End, PageDown, ArrowUp, ArrowLeft, ArrowDown, ArrowRight

Chatter: none detected.
Max simultaneous keys held (rollover observed): 0
What this tool can't see (click to expand)
  • OS-reserved shortcuts never reach the browser at all: Cmd+Tab, Cmd+Space, Cmd+Q on macOS; Win+L, Win+D, Ctrl+Alt+Del on Windows. These are consumed by the operating system before any web page can see them.
  • Some browser-reserved combos are eaten by the browser chrome even in this tool: Ctrl+W (close tab), Ctrl+T (new tab), Ctrl+N (new window), and on Firefox the F6/Ctrl+L address bar shortcuts. Full-screen mode (button above) reclaims some — but not all — of these.
  • Fn-layer behavior (the blue-printed secondary functions on many keyboards, and hardware media keys some laptops remap) lives entirely inside the keyboard's own controller — it never generates a distinguishable browser event, so this tool cannot verify it.
  • NumLock's state via getModifierState is unreliable on some platform/browser combinations — shown as "n/a" in the modifier bar rather than a possibly-wrong on/off.
  • Pressing Esc while full-screen always exits full-screen first (a browser security behavior that can't be overridden) — the first Esc here also counts toward the double-Esc release gesture, so Esc, Esc reliably both exits full-screen and disarms capture.
  • The "Show my layout" overlay depends on the Keyboard Layout API, which only Chromium-based browsers (Chrome, Edge) implement — Firefox and Safari show an honest "not available" label instead of a guessed or broken overlay, and even in Chromium it only reports keys that produce a printable character (not Escape, F-keys, arrows or modifiers).

About this tool

Press any key and it lights up on a visual keyboard (ANSI, ISO or JIS layout; 60%, TKL or full-size), then stays shaded so you can walk the whole board key by key — the exact workflow for checking a new keyboard or verifying a used one. Every press also logs the full event detail — event.key, event.code, legacy keyCode, location, repeat — in a copyable mono log, with a live modifier bar for Ctrl/Shift/Alt/Meta and the CapsLock/NumLock lock states.

Two checks go beyond what most testers offer. Rollover: hold down as many keys as you can and the tool reports how many register simultaneously (your keyboard's real N-key rollover, current and session-max). Chatter: when one physical press registers as two — the classic worn mechanical switch fault — the tool flags any key whose release-to-press gap is under 30 ms, marks it magenta and counts occurrences, so an intermittently double-typing switch shows up with evidence.

The page is honest about what a browser can't see: OS-reserved shortcuts (Cmd+Tab, some Win-key combos) and Fn-layer keys never reach the page, and they're listed as such rather than shown as failures. A full-screen mode reclaims some combos the OS would otherwise swallow, and keys that would scroll or navigate the page are intercepted while testing.

A typing-speed & rhythm mode has you type one short sample sentence and reports WPM alongside the mean, median and standard deviation of the gap between your keystrokes, plus a small histogram of that rhythm — a better feel for a switch or keycap profile than a WPM number alone. Where Chrome/Edge support it, a toggle loads the Keyboard Layout API and overlays the actual character each physical key produces on your OS layout directly on the visual keyboard — genuinely useful once you're off a US layout, where the default QWERTY diagram is misleading; unsupported browsers are told so plainly instead of shown a broken button. And a full stress-test report turns a whole test session into one copyable summary: completion percentage for every key on the selected layout, the exact list of keys you haven't pressed yet, every chatter event detected, and the highest simultaneous-key count you reached — a paste-ready writeup for a marketplace listing or a warranty claim.

Frequently asked questions

A key doesn't light up — is it broken?

First check whether it's on the not-testable list (OS-reserved combos, Fn layer, some media keys) — the browser never receives those, and the page lists them explicitly. If a normal letter/number key shows nothing in the event log either, and the same key fails in another app, the switch or membrane is genuinely dead.

What is key chatter and how does the detector work?

Chatter is one physical keypress registering as multiple presses — a worn or dirty mechanical switch bouncing electrically. The detector flags any key whose keyup→keydown gap is under 30 ms, far faster than a human double-tap. If a key accumulates chatter events during normal typing, that switch needs cleaning or replacing.

How many keys should register at once?

Office keyboards commonly support 6-key rollover (6KRO) plus modifiers; gaming keyboards advertise NKRO — every key simultaneously. Hold down as many keys as you can across the board: the max-simultaneous count is your real rollover, and if it caps at 6, that's 6KRO hardware (or a USB mode limit some boards let you switch).

Why don't Cmd+Tab / Win+L / Fn keys show up?

The operating system consumes those shortcuts before the browser sees any event — it's a security/UX boundary, not a keyboard fault. Full-screen mode recovers some combos, and Fn-layer behavior lives entirely inside the keyboard's own controller, invisible to any software test.

What does the typing-speed mode actually measure?

Standard gross WPM — (characters typed / 5) divided by minutes elapsed — plus the rhythm behind that number: the mean, median and standard deviation of the gap between consecutive keystrokes, and a small histogram of that distribution. It's honest about its own limits too: timing is measured as delivered to the browser tab, so it includes any OS/browser input-lag jitter, not just your fingers, and held-down auto-repeat is never counted as a fresh keystroke.

Why show my keyboard's real layout at all — isn't the diagram already right?

The visual keyboard is drawn as a physical reference (US ANSI/ISO/JIS), but what each key actually types depends on your OS keyboard layout setting — on AZERTY, Dvorak, or a non-Latin layout, that can be very different from the printed diagram. The "Show my layout" toggle calls the Keyboard Layout API to overlay the real character per key. It's Chromium-only (Chrome/Edge); other browsers get an honest "not available" label instead of a silently wrong overlay.

What's in the full stress-test report?

One copyable plain-text summary: how many of the current layout's keys you've pressed and what percentage that is, the exact codes of any keys you haven't tested yet, every chatter event detected (which keys, how many times), and the highest number of keys you held down simultaneously (your observed rollover). It's built for pasting into a marketplace listing, a warranty claim, or a forum post about a specific unit.