Interactive diagnostic

Browser performance test

Compare short browser execution and temporary storage measurements without treating them as an exact hardware ranking.

Processed locally in this browser

Controlled three-run series

Measure the browser execution path

Uses the median of three short runs. Each run includes three CPU work windows, six animation frames, and 512 KB of temporary IndexedDB data. It does not produce exact hardware speed or a device ranking.

Only the measurement summary is kept in this browser for 7 days and added to Device Check and the JSON report. No raw frame, cookie value or site content is stored.

About three seconds of measurementsKeep this tab visible and pause heavy work during all three runs

Test guide

What this test proves—and what it cannot

The browser performance test runs short, bounded execution, animation-frame, memory-visibility, and temporary IndexedDB samples. It is designed for repeat comparisons in the same browser and power conditions, without presenting a universal device ranking or claiming to identify a slow physical component.

Last reviewed: 2026-09-15

Signals measured

Median arithmetic work rate and variation across three short runs
The longest gap observed across six requested animation frames
Median write and read rates from a temporary 512 KB IndexedDB value per run
JavaScript heap use and limit only when the browser exposes non-standard memory data

How the test works

CPU work is split into three 45 ms windows with a yielded frame between windows.
Frame gaps are sampled with requestAnimationFrame and are not converted into a display refresh-rate claim.
IndexedDB receives generated bytes, reads them once, deletes the value, closes the database, and requests database deletion.
The saved seven-day result contains summary numbers only; generated bytes and heap contents are never persisted.

Interpreting the results

SignalWhat it meansNext step

Comparable repeated resultThe same browser completed all short samples under the current foreground load.Repeat after one controlled change, using the same power mode and similar tab load.

CPU or frame result variesScheduling, background work, power state, temperature, and browser activity can affect the three-run series.Repeat the full series under the same conditions and compare its median and variation before drawing a conclusion.

IndexedDB cannot be measuredThe API may be blocked, unavailable, busy, or restricted in this browsing mode.Compare normal browsing mode and site-storage policy before clearing any data.

Measurement limits

The CPU number measures one JavaScript workload, not whole-device, native-application, multi-core, or sustained performance.
A frame gap is affected by browser scheduling and does not reveal panel refresh rate, input latency, or GPU frame rate.
IndexedDB throughput uses one small origin-local sample and is not a physical disk benchmark.

Missing or privacy-reduced detail is not scored as a fault by itself.

Questions about this test

Why does the score change between runs?

Short browser samples share resources with tabs, extensions, the operating system, power management, and thermal control. Compare several runs under matched conditions.

Does a low number prove that my CPU or disk is faulty?

No. It describes this browser workload at that moment. Confirm a persistent real-world symptom with operating-system or manufacturer diagnostics before considering hardware service.