Measurement tool

Download Speed Test

A focused download-only test. It runs a sequence of real transfers, ignores the connection ramp-up and shows every request it made — so you can see the evidence behind the number.

Download speed test

Real downloads from the nearest measurement edge, with every transfer listed.

idle

Measured download speed

— Mbps

Sampled from real transfers to the nearest measurement edge.

Peak sample

—Mbps

Data used

—

Duration

—

Throughput samples during the test
Show the individual transfers
TransferSizeTimeSpeed
No transfers yet.

Each row is one real request. The reported result averages the last stable transfers and ignores the TCP ramp-up.

Ready. Nothing is measured until you start.

How the download figure is produced

The test requests data in growing chunks: a small request to warm the connection, then progressively larger ones. For each request it records a byte-by-byte timeline with the browser's Performance timings, skips the first fraction of a second while TCP slow-start is still ramping, and averages the steady portion of the final stable transfers.

Two safeguards matter for honesty. First, the test stops early when two consecutive transfers agree within 8% — there is no point transferring another 50 MB to prove the same number. Second, if a transfer fails the tool says so instead of substituting a number.

What download speed do you actually need?

Requirements depend on what you do at the same time, not on a single headline number. As a practical guide:

ActivityComfortable downloadNotes
Browsing, email, social5–10 MbpsLatency matters more than raw speed here.
HD streaming (1080p)5–8 Mbps per streamMost services adapt, so a slow line means lower quality.
4K streaming25 Mbps per streamAdd headroom for other devices in the house.
Game downloads and updates50 Mbps+Large updates are where a slow line hurts most.
Several devices at once100 Mbps+One 4K stream plus calls plus updates adds up quickly.

These are service requirements, not marketing targets. A stable 50 Mbps line beats a fluctuating 300 Mbps line for everyday use.

When the download number looks wrong

Run the test three times and watch the spread. If results jump around by more than about 20%, the path is unstable rather than merely slow — the connection stability monitor will show that more clearly than a single throughput figure.

If you are on Wi-Fi, repeat the test next to the router and then over Ethernet. A large improvement when wired is the clearest signal that the wireless link, not the ISP, is the limit.

Questions about this tool

Why does the result differ from my ISP plan?
The plan describes the access line. The browser measures the whole path including Wi-Fi, your device and the route to the test edge. Compare a wired test with a wireless one before drawing conclusions.
Does the download test use data?
Yes — it is a real transfer. The amount used is shown in the results panel, typically 20–80 MB depending on how fast your line is.
Why is the first transfer slower than the rest?
Connections ramp up: TCP slow-start, TLS negotiation and DNS lookups all cost time. That is why the first request is ignored when the final figure is calculated.
Can I test download speed on mobile data?
Yes, but check your data allowance first. A full run can consume tens of megabytes, and some mobile operators apply traffic shaping that changes the result.