Next.js speed statistics: 529 websites measured

By Christian Schreiber · Updated October 16, 2026

Next.js is the framework behind many large brand sites and headless shops. It can render pages on the server and ship very little JavaScript, but many sites still hydrate large client-side bundles on every page, and that cost shows up in the field data.

Causes and fixes: Core Web Vitals for Next.js.

By Core Web Vitals pass rate, Next.js ranks fourth of six platforms with their own page, behind WordPress, ahead of Contentful. Clearest advantage over all measured websites: Page weight (median), 4.1 MB against 4.3 MB. Largest gap: JavaScript (median), 2.0 MB against 1.6 MB.

Core Web Vitals on Next.js

41% of the 529 measured Next.js websites pass all three Core Web Vitals with real-user data. Across all measured websites the figure is 49%.

Metric	Good	Median	Middle half	Threshold
LCP	70%	2027 ms	1598 ms – 2690 ms	2500 ms
INP	62%	177 ms	129 ms – 247 ms	200 ms
CLS	72%	0.06	0.01 – 0.12	0.10

Median mobile Lighthouse score: 37. Why lab score and field data diverge is explained in the Website Speed Statistics.

Who runs Next.js in this sample

Next.js sites in this sample are mostly brand sites, headless shops and software companies, often run by in-house teams and hosted on platforms such as Vercel or behind a CDN. The framework allows very different architectures, from static pages to fully client-rendered apps, and the field data mix all of them.

How to read your own Next.js values

On Next.js, INP tells you most about your own build, because it reflects how much JavaScript is hydrated on the page. Compare it with the faster quarter of this sample rather than the median. LCP and time to first byte depend more on hosting and caching choices, so compare them with sites that render the same way, static or on demand.

Next.js by region

Field data of Next.js websites per region.

Region	Websites	Pass all three	LCP p50	INP p50	CLS p50
Worldwide	529	41%	2027 ms	177 ms	0.06
DACH	43	65%	1598 ms	180 ms	0.04
United Kingdom	44	39%	1835 ms	186 ms	0.05

From 30 websites per region.

More metrics: TTFB, FCP, TBT and Speed Index

Thresholds and sources: Website Speed Statistics.

Metric	Source	Good	Next.js median	All websites median	Threshold
Time to First Byte	Field data	58%	707 ms	684 ms	800 ms
First Contentful Paint	Field data	66%	1491 ms	1536 ms	1800 ms
Total Blocking Time	Lab	9%	1357 ms	771 ms	200 ms
Speed Index	Lab	4%	8906 ms	8631 ms	3400 ms

Page weight and third-party code

A Next.js home page transfers 4.1 MB at the median, of which 2.0 MB is JavaScript. 96% load third-party code, from a median of five providers.

Provider	Share	Main thread (median)
Google Tag Manager	66%	522 ms
Meta (Facebook)	32%	299 ms
OneTrust (Optanon)	24%	109 ms
Google Ads (DoubleClick)	17%	167 ms
TikTok	13%	176 ms
Other Google APIs and SDKs	8%	39 ms

Most common lab findings

Finding	Share	Saving (median)
Render-blocking resources	90%	2853 ms
Unused JavaScript	83%	1560 ms
Images not properly sized	74%	1102 ms
Unused CSS	47%	300 ms
Images not in next-gen formats	31%	1336 ms

Typical Next.js performance bottlenecks

Hydration of big React trees is the typical reason for a slow INP on Next.js sites. LCP suffers when the hero image is rendered on the client or loaded without priority. Server components, smaller client bundles and the built-in image component with priority on the LCP image address most of it.