Quick test · five trials · device delay removed

Reaction Time Test

Test your reaction time in five trials and see your true reflex with the mouse, browser and screen delay measured and subtracted.

Reaction time test · latency-correctedFive trials · click or press Space · device delay measured and removed
Click to start Wait for green, then click as fast as you can. Clicking early restarts the trial.
Your reactionMedian of five, device delay removed
Raw medianWhat a normal test would report
Device shareClick → frame delay on this setup
Best trial0 of 5 trials

Why two numbers: every reaction test includes the time your mouse, browser and screen take to notice the click. This page measures that delay on each trial and subtracts it, so the number you keep describes you, not the desk. Compare the raw median with other sites; compare the corrected one with yourself.

Keep and share this result

Save the numbers to your private Device Passport in this browser, or copy a card that shows exactly what was measured.

Open passport

Nothing is uploaded unless you contribute. Cards and links contain only the numbers shown above.

Words on this page
Reaction time

How long you take to respond to something appearing. Every published number includes your hardware delay too, unless the test measures and removes it.

Input latency

The delay between your action and the result appearing. What a browser can see is one slice of it: the event arriving and the next frame being drawn.

Refresh rate

How many times per second the display redraws. It caps how quickly anything you do can become visible, whatever the rest of the chain does.

Quick answers before you test

What is a good reaction time?

Around 250 ms is the human average for a simple visual reaction; 200–230 ms is fast, under 180 ms is elite. Most tests inflate the number by 15–40 ms of device delay, which this page measures and removes.

Why is my score worse than on other sites?

Usually it is not. Screen refresh, mouse polling and browser scheduling add time on every setup; a 60 Hz display alone adds up to 16 ms before you can even see green. The raw median here is comparable with other sites; the corrected number is comparable only with yourself.

See where the device delay comes from →

Does a faster mouse help?

A 1000 Hz mouse removes a few milliseconds compared with 125 Hz, and a 144 Hz display removes more. The device-share tile shows exactly how much your chain costs.

What the result actually measures

Each trial waits a random 1.5–4 seconds, paints the box green on an animation frame and records the timestamp of the next click or Space press. It then measures how long the browser takes to deliver that click to the next painted frame, which is the visible part of the device chain, and subtracts the median of those delays from the median reaction. Five trials make a result; clicks before green count as false starts and restart the trial.

Measurement boundary: The correction removes the browser event-to-frame delay it can measure, not the display's own pixel response or the mouse switch debounce, so the corrected value is still an upper bound. Fatigue, caffeine, screen brightness and looking away all change the result; treat one session as one data point.

How to interpret the output

Corrected reaction

Your reflex with the measured device delay removed. Under 180 ms elite, 180–230 fast, 230–280 average, above 280 slow for today.

Raw median

What a plain reaction test reports on this setup. Use it to compare with other sites and friends on their own setups.

Device share

The event-to-frame delay measured while you clicked. Large values point to a 60 Hz display, a low polling rate or a busy browser.

Troubleshooting unexpected results

Every trial says too early

Wait for the box to turn green; the delay is random so it cannot be anticipated. Space presses count as clicks, so keep your hand off the bar while waiting.

Results swing by 50 ms

Normal across five trials. Run three sets and compare medians; a single fast trial is anticipation, not skill.

Device share is very high

Close other tabs, disable browser extensions that inject into pages and check the display refresh rate; a phone in battery saver mode also caps the frame rate.

Frequently asked questions

Is this the same test as human benchmark sites?

The raw number is measured the same way. The difference is the second number: the device delay is measured on every trial and removed, so you can see your own reflex separately from the hardware.

Can I use a controller?

Not yet. Mouse clicks and the Space bar are counted; controller buttons pass through a different path with its own delay.

Why five trials?

Fewer than five makes the median unreliable; more adds fatigue. Run the test again for another five and compare.

Can a chatbot measure my reaction time?

No. It needs a green square on your screen and your click on your device, right now.