Connection test - latency, ping and connection type
Measured now
Reported by your browser
Firefox and Safari do not report these at all, so "Not reported" there is expected and is not a fault. Where they are reported, they are the browser's running estimates from its own recent traffic - rounded coarsely on purpose - rather than anything measured at the moment you pressed the button.
Why there is no download speed here. Measuring throughput honestly means transferring a large file, and over a short transfer you measure the connection ramping up rather than its capacity - so a figure derived from small requests would simply be wrong. Latency is the thing that actually makes a connection feel slow or responsive, it can be measured properly, and it is what this page measures. For a bandwidth figure, use a dedicated speed test.
Related Tools
More tools in the same category
Other Tools
Popular tools from across A2Z
Latency is what makes a connection feel slow
People usually reach for a "speed test" when what they are experiencing is a latency problem. Bandwidth decides how long a large download takes; latency decides how long anything takes to start, and it is what makes a video call stutter, a game feel unresponsive and a page feel sluggish even on a fast line.
This measures latency properly: several small requests to this server, timed, with the median reported so one delayed probe cannot skew the result. The first probe is discarded, because it carries the connection setup - DNS, TCP and TLS - that the later ones reuse, and including it would report a handshake rather than a round trip.
Reading the result
| Latency | What it means |
|---|---|
| Under 40 ms | Excellent. Fast enough for anything, including competitive games. |
| 40 to 100 ms | Good. Video calls and most games will be fine. |
| 100 to 250 ms | Moderate. Noticeable in calls and games; browsing still feels normal. |
| Over 250 ms | Slow. Pages feel sluggish to start, and real-time use is difficult. |
Some latency is simply distance - light takes time, and a server on another continent cannot be reached in 10 ms no matter what you pay. Compare against a server near you before concluding anything about your line.
Jitter matters more than the average for calls
Jitter is how much the latency varies between probes. A steady 90 ms is comfortable for a video call; a figure that swings between 30 and 300 ms is not, even though the average looks better. Variation is what produces the robotic audio and frozen frames people describe as "bad internet", and it is the number to watch if calls are your problem.
High jitter on Wi-Fi is often interference or distance from the router rather than anything your provider controls. Test again on a cable before reporting a fault.
Why there is no download speed here
Measuring throughput honestly means transferring a large file. Over a short transfer you measure the connection ramping up rather than its capacity, so a figure derived from a handful of small requests would simply be wrong - and pushing megabytes on every page view would be a poor trade for a number that dedicated speed tests already provide well. This page measures one thing correctly rather than three things badly.
What your browser reports
The second group of readings comes from the browser's own network information, where it exposes any. These are running estimates maintained from recent traffic, deliberately rounded and capped for privacy reasons, rather than measurements taken when you pressed the button - so treat the downlink figure as a rough class rather than a speed. Firefox and Safari do not report this at all, and "Not reported" there is expected rather than a fault.
Rate this tool
Was this tool useful? Your feedback helps us improve it.