Internet

Detecting 4G or 5G in a browser: what a website can actually know

Pingsivo · 1509 words · Updated 7 October 2026

English editorial adaptation prepared with AI assistance. Hypothetical examples are labelled. Technical sources accompany the relevant explanations.

A high speed result does not prove that a device is using 5G. Browser connection information is limited and varies across platforms; an effective connection category is not a reliable label for the physical mobile generation. Pingsivo therefore uses a declared connection type rather than claiming automatic 5G, premium-plan or network-slice detection.

Speed is not a network fingerprint

Different access technologies can produce overlapping speed results. Conditions, device capabilities, concurrent use and the route all influence a measurement. One throughput value cannot uniquely identify the radio generation.

The same principle applies to low results: a slow observation does not prove that a device has fallen back to an older technology. Establish what the device or network actually reports before attaching a generation label.

Keep measured performance separate from declared context. “Measured 80 Mbit/s; user reports mobile data” is clearer than an unsupported automatic “5G detected” badge.

Understand effectiveType

The browser's Network Information API may expose an effective connection category on supported platforms. The MDN effectiveType documentation describes this as an estimate related to observed connection characteristics, not a reliable inventory of the radio technology.

A category named 4g should not be turned into proof of a physical 4G connection. It may be present in contexts where the underlying access path is different. Availability also varies, so missing information is not a diagnostic failure.

Read the broader Network Information API documentation before designing an interface around these values. Show uncertainty rather than guessing where the browser offers no reliable signal.

Separate mobile access from the browser's local path

A laptop tethered to a phone can use Wi-Fi locally while the phone uses a mobile network upstream. One label cannot fully describe both parts of that path.

Similarly, a phone may be connected to a home Wi-Fi network even though its status bar shows mobile service availability. Check the device's actual configuration rather than assuming that the visible radio icon identifies the path used by every browser request.

A useful report records both local connection and upstream context when known. The Wi-Fi and Ethernet guide explains why layered paths matter for interpretation.

Use declarations honestly

Pingsivo lets users describe their connection. This helps organise comparisons, but a declaration is not independently verified detection. Keep that distinction in the interface, exports and shared images.

Choose labels that describe what you know. If you are unsure, use unknown rather than inventing a generation. A careful unknown is more useful than a precise-looking but unsupported value.

When sharing results, state how the information was obtained: device setting, user declaration or an actual measurement. These are different evidence types and should not silently replace one another.

Be especially cautious about premium 5G claims

A consumer speed test cannot establish a subscription entitlement or prove that traffic used a particular network slice merely because the result looks fast. Such claims require suitable operator or platform information and an explicit measurement scope.

Do not infer a premium plan from an IP address, a provider name or throughput alone. Those clues cannot establish all the details of the user's commercial and network configuration.

For Pingsivo, the useful goal is explaining the observed experience and its limits. Adding an impressive badge without reliable evidence would reduce trust rather than improve the diagnosis.

Organise a useful mobile comparison

Keep device, position, tool and profile consistent. Record time, declared access context, movement and whether tethering is involved. Mobile conditions can change, so repeat the scenario rather than keeping a single favourable result.

Read data-use information before choosing long or repeated tests. Three throughput runs consume more data than one. A short stability observation answers a different question and does not replace a capacity measurement.

Use the reliable-test guide and speed-unit guide to keep comparisons meaningful. If the device changes generation during the series and you can observe that reliably, record it rather than hiding the change.

Share the result with its limits

Include the measured values, destination, method, time, completion status and declared context. Avoid claiming that one mobile result predicts every place or time.

A hypothetical fast result in one location is evidence about that observation, not a coverage map or a promise of premium service. Repeat relevant scenarios before deciding on equipment or an offer.

For intermittent behaviour, the stability guide can add context while retaining its HTTP-specific limitations.

Design a mobile test record without pretending to detect radio technology

A useful record separates three categories: what the user declares, what the device explicitly reports through a reliable interface, and what the speed test measures. These fields should not silently overwrite one another. If the user selects mobile data, store that as a declaration. If the browser provides an effective category, describe its actual meaning and support limitations. Store throughput and response timing as measurements towards the test destination. This structure prevents a later export or dashboard from turning a contextual label into an unsupported technical finding.

Consider a hypothetical laptop connected to a phone's hotspot. The laptop's immediate link may be Wi-Fi while the phone's upstream service uses a mobile network. A result labelled only Wi-Fi would omit part of the context, while a result labelled detected 5G would claim more than the laptop's browser established. A clear description can say that the user reports tethering through a phone and can separately record any radio information checked on that phone. The measurement still describes the combined path, including the local tethering link and the route to the destination.

For a direct phone comparison, record whether Wi-Fi is enabled and which path the device is actually using according to its settings. Do not infer this solely from a decorative status icon or from the speed result. If you cannot establish the path reliably, mark it unknown. That is not a failure of the report; it is an accurate statement about available evidence. The rest of the measurements can still be useful for comparing the experience at that location, provided the uncertainty remains visible when the result is shared.

Repeated mobile observations need context about movement and timing. A series collected while stationary in one place is different from a sequence collected while travelling. Record the scenario without collecting unnecessary precise location information. If the question concerns a working location, repeated stationary comparisons may be more interpretable than a moving test. If the problem happens during travel, describe that use case explicitly and recognise that a short browser observation does not identify every handover or network event along the journey.

Data consumption is part of the test design. Read the selected profile's approximate payload and consider the cost of repeated runs on the connection you are using. A more demanding profile is not automatically a better answer to every question. Throughput and short stability observations serve different purposes. Choose the method that addresses the problem, and retain partial results if you stop a test rather than pretending the run completed. This makes the record useful without encouraging unnecessary data transfer just to produce a larger number.

Claims about a premium plan or network slice need a separate source of truth. A website should not promote a guessed entitlement because the result crosses an arbitrary throughput threshold. Even when an operator name can be inferred from an address-related service, that information does not establish the user's commercial offer or the treatment of a particular flow. Keep product language modest and explicit. A trustworthy interface explains the measured experience and offers useful next actions without pretending to inspect network controls that it cannot reliably observe.

When preparing a comparison for support, include the declared setup, time, destination, profile and actual application symptom. Distinguish a device-observed radio label from a browser-derived effective category if both are included. Do not combine them into a single verified generation field. This preserves the evidence needed for a more specialised investigation and avoids misleading readers who receive only the exported summary. Honest uncertainty is particularly important in mobile testing because an attractive label can otherwise look like a capability the program does not actually have.

FAQ: mobile detection limits

Does a high speed prove 5G?

No. Performance ranges overlap across technologies and depend on conditions. Use reliable device or network information for the access context, and keep that separate from the measured speed.

Why can effectiveType say 4g on another connection?

It is an effective category based on connection characteristics, not a guaranteed physical radio-generation detector. The label must be explained in that context.

Can Pingsivo recognise my premium plan?

No. The current test does not verify operator entitlements or network slices. User-declared context remains explicitly declared.

Does tethering invalidate the result?

It changes what the result describes. The measurement includes the tethering path and upstream connection. Record both rather than presenting it as a direct mobile-radio-only test.

Should I force a network generation in settings?

Do not change settings merely to obtain a desired label. If a documented, authorised comparison requires a setting change, record it and follow the device or operator guidance.

Related guides and next steps