LumaProbe sends several small HTTP requests and measures the round-trip time. This works in a browser without special permissions, although it is not the same protocol as command-line ICMP ping.
Distance to the UK server can raise latency outside nearby regions.Uses about 24 MB per completed test.
Run several tests at different times for a representative result.
These are approximate single-device estimates, not guarantees. Other users, applications, routes and service requirements can change the experience. How the estimates are interpreted →
Lower, consistent latency generally feels more responsive. Distance to the test server matters, so compare results using the same test and location. Jitter is especially relevant to calls and games.
LumaProbe sends ten small HTTPS requests to its own endpoint and times each complete browser round trip. Idle latency is the median of successful samples; jitter is their standard deviation. During the download it samples latency again to show responsiveness under load. This is not an ICMP command-line ping, so TLS, HTTP processing and browser scheduling are included.
That extra overhead makes the number unsuitable for direct comparison with every game’s in-client ping. It remains useful for before-and-after comparisons made with the same device, browser and destination.
Use one device and one browser for comparisons. Pause cloud backups, game downloads, streaming and large file transfers, then note whether the device is connected by Wi-Fi, Ethernet, mobile data or a VPN. Run the check at least three times instead of choosing a single unusually good or bad result.
Change only one condition at a time. For a Wi-Fi problem, compare beside the router with the affected room. For a suspected broadband fault, add an Ethernet result if practical. Record the time because congestion and provider work can make the same connection behave differently during the day.
Lower and more consistent values generally support calls, remote desktops and games. The appropriate level depends on distance and application: variation and repeated spikes can be more disruptive than a stable average. A large upload or download can create queueing even when the idle result is good.
If latency rises only while the connection is busy, investigate bufferbloat or uncontrolled background traffic. If it is high when idle on Ethernet, compare another destination and check the provider’s status. If only Wi-Fi is affected, focus on the local wireless link.
A browser sees the complete path from this device to LumaProbe, including the device, local network, provider route and test server. It cannot isolate the broadband line on its own. Browser scheduling, security software, VPNs, extensions and background activity can also influence a short test.
Treat the result as diagnostic evidence, not a contractual line measurement or a live provider-outage declaration. A provider can test a different network boundary and report a different valid number. The most useful conclusion usually comes from controlled comparisons and repeated observations.
Retest with large transfers paused and a VPN disabled temporarily if safe. Compare Ethernet with Wi-Fi. For games, also use the game’s own regional latency display because its servers and route are different from LumaProbe.
Do not buy “gaming” equipment from one ping result. First establish whether queueing, Wi-Fi coverage, a particular route or the access connection is responsible. Smart Queue Management can help busy-line latency, but only on compatible, correctly configured hardware.
Read the complete LumaProbe test methodology and privacy notes →
No. It times small HTTPS requests, so browser, TLS and HTTP processing are included. It remains useful for repeated comparisons to the same destination.
Lower and consistent is generally better, but distance and application matter. Compare jitter and loaded latency as well as the idle number.
The game uses another server, route and measurement method. Use its regional latency alongside LumaProbe rather than expecting identical numbers.
Some links may earn us a commission at no extra cost to you. Recommendations are categories, not paid rankings.
Affiliate disclosure