Core Web Vitals Benchmarks
LCP, CLS, FCP, TTFB, TBT and Speed Index percentiles measured across audited websites.
One entry per metric, with p50/p75/p90 and Google’s good and poor thresholds.
Download the JSONWhat is in it
| Field | Value |
|---|---|
| lcp · p50 | 2518 |
| lcp · p75 | 4382 |
| lcp · p90 | 8027 |
| lcp · unit | ms |
| lcp · goodThreshold | 2500 |
| lcp · poorThreshold | 4000 |
| cls · p50 | 0.001 |
| cls · p75 | 0.050 |
| cls · p90 | 0.140 |
| cls · unit | score |
| cls · goodThreshold | 0.1 |
| cls · poorThreshold | 0.25 |
| fcp · p50 | 1866 |
| fcp · p75 | 2743 |
| fcp · unit | ms |
| fcp · goodThreshold | 1800 |
| fcp · poorThreshold | 3000 |
| ttfb · p50 | 601 |
| ttfb · p75 | 1004 |
| ttfb · unit | ms |
| ttfb · goodThreshold | 800 |
| ttfb · poorThreshold | 1800 |
| tbt · p50 | 143 |
| tbt · unit | ms |
| tbt · goodThreshold | 200 |
| tbt · poorThreshold | 600 |
| speedIndex · p50 | 5256 |
| speedIndex · unit | ms |
Where the numbers come from
Every value is aggregated from audits SlapMyWeb ran against live public URLs — each one fetched, rendered in a headless browser and measured. No survey, no panel, no modelling. Individual URLs are never published; only the aggregate is, which is what makes the set safe to release openly. The sample is sites people chose to audit, so it is not a random sample of the web — treat it as a benchmark of the kind of site that gets audited, not of the web as a whole.
Using it
CC-BY-4.0: use it anywhere, including commercially and for model training, as long as you credit the source.
SlapMyWeb. “Core Web Vitals Benchmarks.” SlapMyWeb Open Data, https://slapmyweb.com/data/cwv-benchmarks. CC-BY-4.0.