How Browsers Measure Wi-Fi Quality Without Hardware Access
This tester compares the stability of HTTPS requests from your browser to a Cloudflare edge endpoint. It cannot measure physical Wi-Fi signal strength, antenna RSSI in dBm or signal-to-noise ratio. A poor result identifies an inconsistent internet path worth investigating; it does not identify which part of that path caused the problem.
After you select Start Live Audit, a zero-byte GET probe begins approximately every 600 milliseconds. Only one request is active at a time, so a slow response reduces the sampling rate. Each request uses no-store caching and omits credentials and the page referrer. Timing includes browser scheduling, connection setup when needed, transport and the destination response. This is a continuous browser latency test, not terminal ICMP ping.
The latest latency is the most recent successful request duration. Average latency uses successful requests in the rolling window. Jitter is the mean absolute difference between consecutive successful requests, with failures and excluded slow readings breaking the pair. Successful readings at or above 1,000 ms are excluded from the raw average and jitter but remain visible in the chart and count toward the scoring disruption rate. Timing alone cannot distinguish tab sleep from real network delay. Although people search for network jitter variance, this displayed value is measured in milliseconds and is not statistical variance. Requests exceeding 1,500 milliseconds count as timeouts; other failures are reported separately. Neither is a direct packet-loss measurement.
How to Compare Rooms
- Start live audit. Connect to the Wi-Fi you want to compare, stay beside the router, and start the test. Allow the rolling window to fill.
- Walk around your space. Keep this tab visible and pause at each location. Use the same device and settings; optional audio makes changes easier to notice.
- Investigate repeatable changes. Compare latency, jitter and failed requests, then repeat the route. Treat poor areas as candidates for investigation rather than confirmed radio dead zones.
For a useful comparison, spend at least 20–30 seconds in each room and longer when requests are slow. Keep device orientation and background downloads consistent. Write down the location and observed values before moving on: the rolling window gradually replaces old measurements. The tool does not record your location or build a floor-plan map. Returning to the starting point helps distinguish a location effect from changing internet conditions.
Wi-Fi Stability Tiers & What They Mean for Gaming, Calls & Streaming
These labels describe this tool's heuristic score. Jitter ranges below are illustrative comparison guides, not score boundaries: average latency and failures can lower the score independently. Even an excellent score cannot guarantee gaming performance, call quality or enough bandwidth for 4K streaming.
| Quality tier | Stability score | Jitter context | How to interpret it |
|---|---|---|---|
| Excellent | 90–100% | Under 5 ms is low variation | Consistent requests on this path; test your actual application too. |
| Good | 70–89% | 5–15 ms is modest variation | Compare with your baseline and check for occasional failed requests. |
| Fair | 50–69% | 15–40 ms is increased variation | Look for repeatable spikes; real-time calls may be more sensitive. |
| Poor | Below 50% | Above 40 ms or repeated failures merits investigation | Possible connection problem, not proof of a wireless dead zone. |
For example, a settled window with median latency of 40 ms, median adjacent jitter of 10 ms and no disruptions has a target score of 100. One timeout among 30 probes gives a 3.3% disruption rate and a target of 95; the display moves toward that target gradually. The rating label only changes after crossing a tier boundary by 3 points, so small boundary changes do not flicker between labels. A high base latency to a distant endpoint can reduce the score even over good Wi-Fi. Compare measurements on the same device and route.
Practical Tips to Fix Wi-Fi Dead Zones in Your Home or Office
Start with router placement: an open, elevated position away from dense obstacles is worth comparing with a router hidden behind furniture. Re-run the same route after each change. The relative behavior of 2.4 GHz vs 5 GHz depends on building materials, interference, channel width and the devices involved; 2.4 GHz often reaches farther, while 5 GHz may offer more capacity at shorter distances. This page cannot detect your current band or choose a channel.
Use router diagnostics or a native analyzer to review congestion before changing channels. If weak coverage repeats in the same room, compare an additional access point or a suitably placed mesh node. Test its backhaul as well as the client connection. If delays rise mainly during large uploads, queueing or bufferbloat may be involved. This idle probe loop does not deliberately saturate the link, so it cannot independently diagnose bufferbloat. An Ethernet comparison can help separate wireless effects from ISP or destination problems.
Frequently Asked Questions
Can a website measure my router's dBm signal strength?
No. This website cannot read antenna RSSI in dBm, Wi-Fi channels or signal-to-noise ratio. Its bars visualize a request-stability score. Use your router diagnostics or a supported native Wi-Fi analyzer for radio measurements.
Why is my Wi-Fi strong next to the router but slow in the bedroom?
Walls, distance, interference and access-point placement can affect a wireless connection. However, internet congestion, VPN routing, browser load and endpoint behavior can also increase these timings. Repeat the same route and compare with a stationary baseline before attributing the change to Wi-Fi coverage.
Does this test use my mobile data?
Yes, if your device uses a mobile connection or switches to one. Each probe requests a zero-byte response body, but request headers, encryption and transport still consume data. There is no fixed total because the test runs until paused or the tab becomes hidden. Disable mobile fallback in device settings if appropriate.
Is the drop rate the same as packet loss?
No. The timeout rate counts HTTPS requests that exceed 1,500 milliseconds. The failed-request rate also includes HTTP and network errors. Browsers do not expose the underlying packet retransmissions here, so neither percentage measures IP or ICMP packet loss.
How is the stability score calculated?
After at least 10 probes and five usable successes, a target score uses median latency and median adjacent jitter from up to 30 probes. Latency above 60 ms costs 0.25 points per ms (maximum 25); jitter above 25 ms costs 0.6 points per ms (maximum 30). Failed requests and successful readings at or above 1,000 ms cost 1.5 points per percentage point of the window (maximum 60). The target is clamped to 5–100. The displayed score moves 20% toward each target, at most 3 points per probe. Rating labels use a 3-point buffer around tier boundaries to avoid flicker. Raw average latency, mean jitter, failures and the chart remain immediate. Hidden tabs pause and clear the session. This heuristic compares an internet path, not radio signal strength.
Continue Checking Your Network
Technical references: MDN Network Information API and Cloudflare speed-test measurements. Browser connection estimates are optional and do not reveal Wi-Fi signal strength.