Astro Passes Core Web Vitals Far More Often Than Next.js

Next.js and Nuxt are marketed on speed: server rendering, static generation, image optimization, all aimed at Core Web Vitals. Astro made a different bet, shipping mostly static HTML and hydrating only the interactive pieces. Across 490,000 sites, the field data says Astro's bet paid off. Astro passes Google's Core Web Vitals on 48% of mobile sites, more than 1.6 times Next.js at 29% and roughly double Nuxt at 25%.

That difference decides real outcomes. Speed determines whether visitors stay, whether they convert, and how a page ranks in Google. Core Web Vitals measure the experience directly: how fast the main content loads, how quickly the page answers a tap, and how steady the layout stays. A meta-framework that passes half as often as another is, in practice, delivering a slower experience to real people on real phones.

These are Google's real-user field readings, the Chrome UX Report collected from actual Chrome visitors, not a lab score on fast hardware, where all of these frameworks look healthy.

We detected the meta-framework on each site from HTTP Archive's technology crawl and joined it to the Chrome UX Report, Google's field record of Core Web Vitals from real Chrome users. That gave 490,000 mobile origins running Next.js, Nuxt or Astro with a field reading. A site passes Core Web Vitals when it is good on all three metrics at once at the 75th percentile of its visitors: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, and Cumulative Layout Shift under 0.1. Measured on mobile, it is the same demanding standard Google uses to judge a page, so a passing framework is genuinely fast for most of its visitors.

On mobile, Astro sites pass Core Web Vitals 48.0% of the time. Next.js sites pass 29.3%, Nuxt sites 24.9%. That is not a rounding difference; Astro's real-world pass rate is more than 1.6 times Next.js's and roughly double Nuxt's.

It also puts Astro above the roughly 42% mobile web-wide average, while Next.js and Nuxt sit well below it. The newest entrant is the only one of the three that makes a site more likely to pass than the average website, and the reason is baked into its architecture: it sends the browser less to do.

As with UI frameworks generally, the decisive metric is responsiveness. Astro sites are good on Interaction to Next Paint 88.9% of the time. Next.js sites only 66.3%, Nuxt sites 72.8%, both below the 82.5% web-wide average.

Next.js and Nuxt default to sending a JavaScript representation of the page to the browser and hydrating it, re-attaching interactivity to server-rendered markup. On a fast machine that is invisible. On a mid-range Android the phone has to download, parse, and execute a substantial bundle before taps feel responsive. Astro's islands architecture inverts the default: the page ships as static HTML, and only the components that genuinely need interactivity receive JavaScript. That is a 22.6-point good-responsiveness advantage over Next.js on mobile, and loading follows the same logic, with Astro good on loading for 76.7% of sites against Next.js's 54.8% and Nuxt's 48.4%.

Next.js is mid-transition between its older Pages Router and the newer App Router built on React Server Components, which are designed to ship less client JavaScript. The data shows the App Router helping at the margin: 31.1% mobile pass versus 29.3% across all Next.js sites, and a clearer desktop lead of 55.8% to 52.4%. It is a real improvement, but it does not close the gap to Astro's static-first baseline.

On desktop the three converge entirely, with good-responsiveness rates between 95.7% and 97.0%. A desktop CPU chews through hydration fast enough that the architectural difference barely registers. That is the single most important caveat for anyone reading a green Lighthouse score on a laptop: it says almost nothing about how the same Next.js or Nuxt app feels on the phone in a customer's hand.

SvelteKit did not reach the 30,000-origin reporting threshold this crawl. Buckets are non-exclusive.

If Core Web Vitals matter to you, this data points at architecture before optimization. Astro passes on 48.0% of mobile sites, Next.js on 29.3%, Nuxt on 24.9%, a 19-point spread that comes almost entirely from responsiveness: Astro is good-INP on 88.9% of sites against Next.js's 66.3%, a 22.6-point lead. The static-HTML-plus-islands model reaches passing scores in the wild far more often than full-hydration React or Vue does.

Read your Lighthouse scores on the right device. On desktop all three converge to between 95.7% and 97.0% good INP, because a laptop CPU absorbs hydration the way a mid-range Android cannot. A green score on your own machine says almost nothing about the phone in a customer's hand, so measure Interaction to Next Paint on real mobile hardware.

If you are staying on Next.js or Nuxt, the gap is not fixed in stone. The App Router already lifts Next.js from 29.3% to 31.1% on mobile, so adopt the lighter-rendering paths your framework offers, partial hydration and streaming among them, and treat client-side JavaScript as a budget you spend rather than a default you inherit. The React vs Vue vs Angular field study shows the same rule at the UI-framework level: the phone rewards whatever sends it less work.