WiFi Speed Test

Testing a wireless connection

The single most useful thing you can do here is run this test over Wi-Fi, then run it again with a cable. If wired is fast and wireless is slow, the problem is inside your home rather than with your connection — and that narrows the fix to the router, its placement, or the band you are on.

How the download figure is produced

The page fetches files of known size from this server with caching disabled, times how long each takes, and converts that into megabits per second. It runs three times and reports the best, because the first request also carries connection setup and TLS negotiation.

What this measures, and against what

The measurement is between your device and this site's server, over whatever route the internet chooses today. That makes it a real measurement, but it is not the same as a dedicated speed-test service, which runs servers close to most users specifically to measure the line rather than the route.

If this reads lower than a dedicated test, the difference is usually distance and routing rather than your connection.

Why the best rather than the average

Because anything that slows one run down — a packet loss event, another device waking up, the browser doing something else — makes the figure too low rather than too high. The fastest run is the one least contaminated, so it is the closest to what the connection can actually do.

How to use it

1. Close other tabs, downloads and calls.
2. Press start and leave it alone for a few seconds.
3. Compare wired against wireless if the figure disappoints — that one comparison isolates most home network problems.
4. Run it again later. One reading is not a measurement.

Common uses and tips

Getting a reading worth trusting

• Close everything else first. A cloud backup or a video call will share the line and halve the figure.
• Use a cable if you can. If wired is fast and wireless is slow, the problem is your Wi-Fi rather than your connection.
• Run it several times. Networks vary minute to minute, and one reading tells you very little.
• Test at different times. Evening congestion is real and is the commonest cause of "it's only slow sometimes".

Why your speed never matches what you pay for

• Protocol overhead takes roughly 5–10 % before anything else.
• Wi-Fi is shared and lossy. A 500 Mbps line routinely delivers 200 over wireless, and much less through walls.
• Old hardware. A router or network card from the 100 Mbps era caps everything behind it, regardless of the line.
• The advertised figure is usually a maximum rather than a guarantee.

Megabits and megabytes

Connections are sold in megabits per second and downloads are shown in megabytes per second. Eight bits make a byte, so a 100 Mbps line downloads at about 12.5 MB/s at best. Both figures are shown above so the difference is visible.

Questions people ask

Why is this slower than a dedicated speed test?

Because it measures the route to this site's server rather than to one placed close to you. Distance and routing account for most of the difference.

Why does it not match my advertised speed?

Protocol overhead, Wi-Fi losses and older hardware all take a share, and the advertised figure is usually a maximum rather than a guarantee.

Why does it report the best of three runs?

Because interference only ever makes a run slower. The fastest run is the least contaminated.

What is the difference between Mbps and MB/s?

Eight bits to a byte. A 100 Mbps line downloads at about 12.5 MB/s at best. Both are shown.