Skip to content

Performance

BeeUI measures repeatable component operations and package/bundle footprint so regressions can be detected against a controlled baseline. Benchmark classes include representative Table work, overlay open/close/positioning, theme/token operations and package footprint.

A benchmark is meaningful only with its environment, warmup/sampling method and variance. CI/regression budgets catch material changes; they are not universal frame-rate or device latency guarantees. Browser bundle size is not the same as native binary size, and synthetic component loops are not end-user interaction traces.

Use BeeUI’s results to compare changes on the same harness, then measure your own app on the hardware/browser mix that matters to you.

Sources: benchmark harness, performance baseline, and bundle footprint.