Speedlify — the Eleventy community leaderboard
A continuously-running measurement of sites built with Eleventy, drawn from the community "Built Awesome" collection. Each site is re-tested once a week. Read 2026-09-02.
What it reports
- 1,145 of 1,232 sites measured, most recent scan 3 hours before reading.
- Median Lighthouse scores: 98 performance, 100 accessibility, 100 best practices, 100 SEO.
- 150 of 1,145 sites hold a clean sweep — 100 in all four categories with no axe accessibility violations.
- Measured with Lighthouse lab runs, axe for accessibility, and Chrome UX Report field data where available.
- Ranking is "circle color first: all green beats a row with amber, which beats red," with Lighthouse category sums and Core Web Vitals as tiebreakers.
A median Lighthouse performance of 98 across eleven hundred independently-built sites is the strongest quantitative support this spoke holds for the claim it has been making qualitatively since static-site-generator was written: that pre-rendered output starts fast and stays fast without anyone tuning it. It is consistent with eleventy-performance-docs — these are the scores of a generator whose page-speed story is a default.
The number that limits it
Only 118 of the tracked sites have field data at all. Roughly one in ten.
That is the whole methodological problem with putting Eleventy on this spoke's platform scoreboard. core-web-vitals quotes Astro at 66–67% of real sites passing and WordPress at 48–49%, and those are CrUX field numbers — real visits by real people. CrUX only reports a site once it has enough traffic to be statistically usable. Nine in ten of the Eleventy sites on this leaderboard never clear that bar.
So the 98 is a lab median, and it is not the same kind of number as Astro's 66%. Lighthouse runs on a simulated device on a simulated connection; Core Web Vitals as Google uses them are field data. A site can score 98 in the lab and have no field record whatsoever, and most of these do.
And the sample selects itself twice over. These are sites their authors submitted to a showcase, so they are the ones people are proud of — and they are small enough not to appear in CrUX, which is a different population from the commercial sites that make up Astro's and WordPress's field numbers.
Why that matters beyond Eleventy
The platform scoreboard on core-web-vitals is only defined for platforms whose sites are big enough to be measured. That systematically excludes the small-site end of the genre — personal sites, docs, blogs — which is exactly where Eleventy, Jekyll and Zola live, and exactly what the static-generation argument is usually made about. The scoreboard is not neutral about site size, and nothing in this spoke had said so.
T2, not T1: real repeated measurement with a stated method, on a self-selected sample, by the ecosystem it measures.
Related
eleventy · eleventy-performance-docs · core-web-vitals · core-web-vitals-wordpress-astro · astro · static-site-generator · jekyll · zola