How Volt measures your speed

Updated 23 September 2026

A speed test is not an absolute truth: it is a measurement, with a protocol and limits. Here is exactly what Volt does, so you know what the numbers on screen are worth.

What the test measures

Measurement	What it represents	Unit
Download	How fast data reaches you	Mbps
Upload	How fast your data leaves	Mbps
Ping	Round-trip time to the server	milliseconds

The servers: M-Lab

Volt runs on M-Lab (Measurement Lab), an open measurement platform supported by universities and non-profit organisations. Its servers sit around the world and their results are published openly.

Two practical consequences:

The server is chosen automatically from those nearest to you. You do not pick it in the interface.
There is no fallback server. If no M-Lab server can be reached, the test fails rather than showing a made-up number. A test that refuses to start beats a test that lies.

The protocol: ndt7

The measurement uses ndt7, M-Lab's test protocol. It opens a connection and sends real data, not a calculated estimate.

Download: 10 seconds. The browser receives a continuous stream and measures what actually arrives.
Upload: 10 seconds. Your browser sends the data and the server measures what it receives, then reports back. That is more accurate than measuring in the browser, which cannot know what really reached the far end.
Ping comes from the TCP connection's MinRTT: the shortest round trip observed during the test. That value reflects the latency of the link without one-off slowdowns.

Ten seconds is a deliberate compromise: long enough to get past a connection's ramp-up phase, short enough not to eat gigabytes of a mobile data plan.

What it consumes

A full test transfers real data: roughly your speed multiplied by the duration. On a 500 Mbps line, a 10-second download test moves several hundred megabytes. On 4G or 5G with a capped plan, be careful: a handful of tests can dent an allowance.

Why your result varies

The same plan never gives the same number twice. The main causes:

The radio link: over Wi-Fi, distance, walls and neighbouring networks change everything. An Ethernet cable removes that variable.
Other devices in your home: a console downloading in the background takes a share of the line.
Time of day: networks are busier in the evening.
Your browser and computer: an older machine can cap out before the line does, especially above 500 Mbps.
The network path to the measurement server, which is not the path to any given service.

To compare seriously, measure in the same spot, at the same time, with the same device, and change one thing at a time.

What the test does not tell you

It does not measure speed towards a specific service. A video that stutters may be the service's fault, not your line's.
It does not replace an official measurement in a dispute with a provider. Several national regulators offer certified tools for that.
It does not test stability over time: ten seconds cannot see an outage that happens three times a day.

Privacy

Volt needs no account. The test runs between your browser and the measurement server. Because M-Lab publishes its measurements openly, technical details of the test, including the IP address used, may appear in that public data: that is how this open platform works, and it is what lets anyone verify the measurements.

Test my speed now

Frequently asked questions

Why can't I choose the server?

The server is picked automatically from the nearest ones, so the measurement reflects your line's quality rather than your choice. To test a specific service, measure from that service instead.

Why does the test last 10 seconds?

Long enough to get past a connection's ramp-up, short enough to limit the data used. A longer run would smooth out variation slightly, but would cost a lot of mobile data.

My result is below my plan: is the test wrong?

Not necessarily. The gap usually comes from Wi-Fi, your device or other activity. Measure again over a cable with nothing else running before drawing conclusions.