HTTP protocol

The HTTP protocol of the HTML document response.

Field data PhoneDesktopAll Scope All sites Q1 2026 edition · Desktop field outcomes
Metric LCP INP CLS
1

At a glance the headline numbers for HTTP protocol

The HTTP protocol of the HTML document response.

3
Categories
94,906 sites
63.1%
Top share
h2
1.4s
Top LCP
Median for h2

23.9% of HTML documents arrive over HTTP/3.

The State of Web Vitals · Q1 2026 · 94,910 sites · desktop field datacorewebvitals.io/state-of-cwv
2

HTTP protocol who uses what, and how fast each group loads

Median LCP
0
0.5s
1s
1.5s
2s
2.5s
3s
3.5s
4s
http/1.11.3s13% of sites
h21.4s63% of sites
h31.4s24% of sites
VariantShare of sitesMedian
http/1.1 12,368 sites
13%
1.3s
h2 59,859 sites
63%
1.4s
h3 22,679 sites
24%
1.4s

Little daylight between the groups: every variant sits near 1.3s median LCP. computed

The State of Web Vitals · Q1 2026 · 94,910 sites · desktop field datacorewebvitals.io/state-of-cwv
3

Passing LCP per HTTP protocol which group passes the LCP most often

VariantSitesPassing LCPLCP
H3 23.9% 22,679 88% 1.4s
H2 63.1% 59,859 86% 1.4s
Http/1.1 13% 12,368 86% 1.3s
Good Needs Improvement Poor Sorted best-passing first · median colored by its own rating · pass = good LCP (2.5s at p75) · one value per site

No variant stands out: pass rates sit between 86% and 88%. computed

The State of Web Vitals · Q1 2026 · 94,910 sites · desktop field datacorewebvitals.io/state-of-cwv
4

Why this matters for the Core Web Vitals, and where to start fixing it

The protocol decides how expensive a connection is. HTTP/3 sets up in fewer round trips and a lost packet no longer stalls everything behind it, which TCP-based protocols suffer from. The document request is where this matters most: it is the first request of the visit and it carries the whole TTFB.

There is a catch on the very first visit. Browsers often discover HTTP/3 support through a header on a previous response (Alt-Svc), so the first document fetch can still pay the older protocol's price. DNS HTTPS records fix that discovery, and your CDN controls whether they exist.

How does this affect the Core Web Vitals?

The choice barely moves the LCP: 88% pass at best, 86% at worst. This signal does not separate passing sites from failing ones.

Related signals Uncompressed text responses → HTML size (kB) → Connection hints → ETag present → Chrome field data from 94,910 sites, representing millions of real page loads · How we measured