Measures how often your browser draws a frame — normally your display's refresh rate — over a few hundred samples, and reports the spread alongside the number. Nothing here flashes, and nothing leaves your device. What it can and cannot see about your actual panel is stated plainly below.
How this test works
The page asks the browser for animation frames and records the time between them. After discarding the first quarter-second — the intervals immediately after an animation loop starts are unrepresentative while the browser settles — it collects up to six hundred samples and takes the median interval. One divided by that interval, in seconds, is the refresh rate. The median rather than the average is the load-bearing choice: a single garbage-collection pause among six hundred frames produces one enormous interval, and an average lets that one frame misreport your display by several hertz. A median ignores it, which is the right treatment for a value that is genuinely constant with occasional interruptions.
Both figures are computed regardless, and the gap between them is what the readout calls jitter. On a clean signal the mean and the median are nearly identical; every stalled frame pushes the mean and leaves the median alone, so the size of that gap is a direct measure of how evenly the browser is delivering frames. The range shows the shortest and longest intervals observed. Frames longer than three times the median are counted and named as stalls rather than folded into the result, because a stall is a fact about your machine at that moment and not a fact about your panel.
Now the limits, which matter more here than on most of these tools. No web page can read your display's refresh rate. There is no API for it, and any page claiming to report your hardware directly is inferring it exactly as this one does. What is measured is the browser's frame delivery, which normally follows the display because the compositor is synchronised to it — and which diverges in several ordinary cases: a background tab, a browser capping frames to save power, a laptop on battery, a variable-refresh-rate panel genuinely changing its rate, or a busy machine. A steady 60 Hz reading on a monitor you believe is 144 Hz is a strong hint the rate is not enabled in your display settings, but it is a hint to go and check, not a verdict on the hardware.
Frequently asked questions
Is this measuring my monitor or my browser?
Strictly, the browser. No web page can read a display's refresh rate directly — there is no API that exposes it. What this measures is how often the browser hands the page an animation frame, which is normally locked to the display because the compositor is synchronised to it. The two agree in the ordinary case and diverge in specific ones: a throttled background tab, a browser capping frame delivery to save power, a laptop on battery, or a variable-refresh-rate panel that is genuinely changing its rate as you watch. The page reports what it can actually observe and names the gap rather than claiming to read your hardware.
Why does it say 59.9 Hz instead of 60?
Because that is what was measured, and rounding it away would discard the only interesting part. Real displays rarely run at exactly their nominal figure — a nominal 60 Hz panel commonly runs at 59.94, a rate inherited from broadcast television. Timer resolution and scheduling noise add a little more. The page shows the measured value, and separately tells you which common panel rate it is consistent with, so you can see both the number and its interpretation instead of only the tidy one.
What do the jitter and range figures mean?
The reported Hz comes from the median frame interval, not the average, because one garbage-collection pause among hundreds of frames would drag an average down and misreport your display. Jitter is the gap between the mean and the median: on a clean signal they are nearly identical, and every stalled frame pushes only the mean. Range is the shortest and longest interval seen. A tight range with near-zero jitter means the browser is delivering frames evenly; a wide range means something on the machine is interrupting them, which is worth knowing in its own right.
Why did it stop when I switched tabs?
Deliberately. Browsers throttle animation frames heavily in background tabs, often to about one per second, to save power. A measurement taken there describes the throttle rather than your screen, so rather than quietly reporting a wrong figure the page stops and tells you why. Start it again with the tab in front. For the same reason, results taken while a video plays or a heavy page renders elsewhere can read low.
Can this tell me if my high refresh rate monitor is set up correctly?
It can tell you what the browser is receiving, which is usually the useful half of that question. If you bought a 144 Hz monitor and this reads a steady 60 Hz, the rate is almost certainly not enabled — check the display settings in your operating system and, on a desktop, that the cable and port support the rate you want. What it cannot do is prove the panel is capable of a rate the browser is not being given, and it cannot distinguish a genuine variable-refresh display from an unstable one. Treat a surprising result as a reason to check the settings, not as a hardware verdict.
Device Bench