DSLReports vs NetSteer — what each test actually shows
DSLReports is a veteran speed-test site with a massive forum database of user results; NetSteer is a free in-browser test that measures latency under real load. They are measuring different things, and knowing which number answers which question saves an evening of guessing. Facts as of 2026-11.
| DSLReports | NetSteer test (this site) | |
|---|---|---|
| Primary metric | Download / upload throughput (Mbps) | Latency added under load (ms) |
| Distinctive data | Thousands of community results for your exact line/ISP — a neighbourhood baseline | Idle vs loaded latency, loss, and a score tied to real-world feel |
| Load model | Single-direction bursts to their servers | Saturated bidirectional load that matches a household in use |
| Bufferbloat | Not measured directly | That is its whole job |
| Cost | Free (ads) | Free, unlimited |
When DSLReports is the better tool
When the question is "is my line actually delivering the speed I pay for, and do people on my street see the same number?". The forum database is the feature — a neighbour on your line showing the same slow result points at the ISP uplink; only you showing it points at your router or wiring.
When NetSteer is the better tool
When the question is "why does my game lag even though my speed test is fine?". That is bufferbloat, and no throughput test will show it. The NetSteer bufferbloat test runs in the browser, takes under a minute, and tells you whether your router is the cause — in which case the fix is SQM/CAKE, not an ISP ticket (see the OpenWrt SQM settings page).
Running them together
One clean hour of diagnosis: DSLReports (ethernet, quiet) for the throughput baseline, then the NetSteer bufferbloat test while the household is actually using the network. If the first is on plan and the second is bad, you have your answer: the line is fine, the router is the queue problem.
More comparisons
Compare NetSteer's test with other popular tools: Waveform · Speedtest.net · DSLReports · Cloudflare
FAQ
Is DSLReports a speed test or a latency test?
It is a speed test with a large community layer: it reports download/upload throughput against their servers and lets you see how nearby users' lines compare — which is genuinely useful for spotting neighbourhood-level problems. It does not, however, measure bufferbloat: it does not report how much extra latency your router adds while the line is saturated, and that number is the one that makes games lag during a download.
Why do DSLReports and my own measurement disagree?
Most of the time it is distance and server selection. DSLReports connects to the nearest of their own test servers; your home network, your ISP routing, and the time of day all shift the number. Repeated, on-ethernet, quiet-hour tests agree better than one-off results. If every DSLReports run is far below your plan speed but a local test (or a neighbour's number in their forums) is near it, the bottleneck is on your side of the wall, not the ISP.
What should I use before an ISP ticket?
Run DSLReports (or any hosted speed test) for the throughput claim, and a latency-under-load test such as the NetSteer bufferbloat test for the lag claim. ISPs distinguish "under-provisioned line" from "router bufferbloat" by exactly that split — the ticket that says "speed fine, latency under load high" often ends at your router, where the fix is SQM/CAKE, not the ISP.