Hardware Tests
Seven free tools for working out whether a piece of hardware is actually broken. They all run entirely in your browser, need no signup, and upload nothing. Each one reads what your computer receives rather than trusting the label on the device, which is usually where the confusion starts.
Keyboard Test
Dead keys, modifier mapping, characters, rollover
Reads the key code each switch actually sends, which is the only thing that settles a Command and Option mix-up. Diagrams for 60%, 75%, tenkeyless and full-size, plus a layout-free mode for anything unusual.
OpenMouse Test
Buttons, wheel, double-click chatter, jitter
Catches the most common way a mouse dies: a worn switch firing twice from one press. It measures the gap between presses and flags anything under 50ms, which no hand can produce.
OpenScreen Test
Dead pixels, refresh rate, banding, ghosting
Nine fullscreen colours for stuck and dead pixels, near-black steps for crushed blacks, and a refresh rate measured from real frame intervals rather than taken on trust.
OpenSound Test
Channels, microphone, frequency sweep
Separates a swapped channel from a dead one, and shows whether your microphone is capturing at all. Record a few seconds and play it back. Every tone starts quiet.
OpenGamepad Test
Buttons, triggers, stick drift, rumble
Measures how far each stick wanders from centre while you are not touching it, which is the drift that makes a character walk on its own. No permission prompt, but the pad stays invisible until you press something on it.
OpenWebcam Test
Real resolution and frame rate, still capture
Shows the resolution and frame rate you actually get rather than the ones on the box, and counts real frames rather than trusting the track. Nothing is recorded or uploaded.
OpenTouchscreen Test
Multi-touch, pen pressure, dead zones
Counts how many fingers actually register at once rather than how many the browser claims, and draws your stroke so a dead patch on the glass shows up as a gap.
OpenWhat these cannot tell you
A web page cannot read your hardware. USB vendor and product identifiers are not exposed to ordinary pages, so none of these tools know what device you have plugged in. They cannot measure mouse DPI, panel response time, speaker frequency response or actual loudness, and anything involving a refresh rate or a report rate is timed from what the browser delivers rather than read from the device.
That sounds like a weakness and is often the opposite. What you usually want to know is not what the hardware is capable of, but what your applications are receiving, because that is the layer a swapped modifier key or a silent microphone actually lives in. I wrote up the specific walls the browser puts in the way in what the platform refuses to measure.