Free Online Keyboard Tester

How to test a keyboard, and what the result means

A keyboard test answers one narrow question well: did the key you pressed arrive in the browser? That sounds modest, but it is the question that splits every keyboard problem into two halves with completely different fixes. If the key arrives, your hardware is doing its job and the fault is further up — in an application, a layout setting, or a remapping tool. If it does not arrive, the signal is being lost somewhere between the keycap and the operating system, and no amount of software troubleshooting will help.

Working through the keyboard

Go row by row rather than hunting for the key you think is broken. It takes about a minute and it catches the keys you rarely use, which are exactly the ones that fail silently for months. Start at Escape, work along the function row, then down through the number row, the three letter rows, and finally the modifiers and the navigation cluster. Each key you press turns from untested to confirmed on the layout above, and the counter tells you how many are left.

Two things are worth doing deliberately. First, hold several keys at once — Shift with a letter, then Ctrl and Alt together, then four or five keys spread across the board. The simultaneous-press counter records the most you managed at any moment, which is a real property of your keyboard rather than of your typing. Second, press each modifier on both sides. Left and right Shift, Control and Alt send different codes, so a keyboard can have a dead right Shift while the left one works perfectly, and you would never notice from normal typing.

Reading the result

Three outcomes, three different next steps.

One key dark, everything else lit. The problem is that key. Debris under the cap is the most common cause and the easiest to rule out — with the machine off, work a thin card around the edge of the keycap, or blow compressed air in at an angle rather than straight down. If it is a mechanical board, swap the switch with a working one from a key you rarely use; if the fault moves with the switch, you have your answer.

A whole row or region dark. The individual keys are almost certainly fine. Keys in a keyboard are wired in a grid, so a single broken row or column trace takes out everything on it at once. On a laptop this often means a partly unseated ribbon cable. On an external board, try a different cable and a different port before concluding anything — a failing cable produces exactly this pattern.

Every key registers here, but the problem persists elsewhere. Your keyboard is fine and the cause is software. Check your active keyboard layout first, since a US board set to a UK or German layout will produce the wrong symbols while every key still registers correctly. Then check for remapping utilities, gaming software with per-profile key bindings, and accessibility features. Sticky Keys and Filter Keys in particular change how presses are interpreted and are easy to enable by accident.

The four faults worth knowing apart

Most keyboard complaints are one of four things, and they get confused with each other constantly because the experience of them is similar.

A dead key sends nothing at all. That is what the test above finds directly. A key with chattersends two presses for one push, so you type "teest" instead of "test" — the switch contacts are bouncing, and the gap between the two signals is usually under 30 milliseconds, far too fast to be you. The double typing test times those gaps and attributes them to specific keys. A stuck key is the opposite: it never reports being released, so it repeats forever or silently modifies everything else you type. A held-down Shift is why a keyboard sometimes appears to have switched to capitals on its own; the stuck key test shows exactly what is still down.

Ghosting is the odd one out, because it is not a fault at all on most keyboards. Press three or four specific keys together on a budget board and one of them stops registering. Nothing is broken — the keyboard is reading a grid and cannot resolve which keys in that grid are pressed without ambiguity, so the controller drops one rather than reporting a key you did not touch. It shows up in gaming far more than in typing, because holding movement keys plus a modifier plus an action key is exactly the pattern that triggers it. The ghosting test finds which combinations fail on your board, and the rollover test tells you the ceiling: two keys, six keys, or unlimited.

What a browser test cannot tell you

Being clear about the limits is more useful than overstating what this does. A page can see the key events the operating system chooses to deliver, and nothing beneath that.

It cannot measure your keyboard's hardware input lag. Event timestamps include the whole path — switch, controller, USB polling, OS input stack, browser event loop — and browsers deliberately coarsen them to prevent timing attacks. Anyone claiming to measure switch response time in a browser to the microsecond is measuring something else. The key latency test explains exactly which parts of that chain it can and cannot see.

It also cannot see keys the operating system intercepts before the browser. Power and lock keys, Fn on most laptops, some fingerprint and media keys, and screenshot combinations on macOS never arrive as key events. If one of those does not light up, that is expected behaviour rather than a fault. And it cannot distinguish a worn-out switch from a dirty contact or a damaged trace — all three look identical from up here, as a key that simply never arrives.

Within those limits, though, the test is decisive about the thing that matters most: whether to keep troubleshooting your software or start pricing a replacement.