Mobile vs desktop Core Web Vitals: INP gap widest, CLS flips

A widely cited figure says 48% of mobile origins and 56% of desktop origins pass all three Core Web Vitals as of July 2025. That headline hides more than it shows, because “passes all three” collapses three very different stories into one number. Broken apart by metric, using the same July 2025 CrUX data behind that figure, the mobile-desktop gap is not one gap at all: it is wide for one metric, narrow for another, and reversed for the third.

INP has the widest gap of the three#

Interaction to Next Paint is where mobile and desktop diverge most. The 2025 Web Almanac’s performance chapter puts desktop INP at 97% good, “maintaining the high standard established in previous years,” against 77% good on mobile, up from 74% in 2024. That is a 20-point gap, and the chapter notes it directly: the mobile-desktop split “has begun to narrow, shrinking from 23 percentage points in 2024 to 20 percentage points in 2025,” but a 20-point gap is still the largest of the three metrics by a wide margin. The chapter does not state a specific cause for the underlying gap itself, only that it is narrowing. Google’s own web.dev guidance on INP is similarly careful: it flags “considerable variability in the capabilities of devices people use” as a reason INP thresholds have to accommodate a wide device range, without pinning the mobile-desktop split on any one factor such as processing power or network conditions.

LCP splits by a more moderate margin#

Largest Contentful Paint passes on 74% of desktop origins against 62% on mobile, a 12-point gap, roughly half the size of the INP split. The Almanac also reports mobile shows “nearly double the rate of ‘poor’ experiences” on LCP compared with desktop, 13% against 7%, so the mobile shortfall is not just fewer origins clearing the “good” bar; more of them land in the worst tier outright. As with INP, the chapter documents the size of the gap without stating what drives it.

CLS is the one metric where mobile wins#

Cumulative Layout Shift reverses the direction of the other two: 81% of mobile pages pass CLS against 72% of desktop pages, a 9-point advantage for mobile. The Almanac notes this is not a one-off snapshot either, stating that “mobile maintains a higher share of good CLS than desktop throughout the entire timeframe” when the chapter looks at the trend from 2023 through 2025. The chapter offers no explanation for why CLS runs opposite to LCP and INP; it only documents that it does, consistently, across two years of monthly data.

Why we care#

The all-three figure hides which metric to fix first. A site told it fails “Core Web Vitals” on mobile could be failing LCP, INP, or both, and the fix for each is different: LCP work generally targets render-blocking resources and image weight, INP work targets main-thread responsiveness to input. Treating “Core Web Vitals” as one problem instead of three risks fixing the wrong one.

INP deserves more attention on mobile than the combined figure suggests. A 20-point gap is the largest disparity in the dataset, larger than LCP’s, and it is the metric most directly tied to how a page responds to a visitor’s taps and clicks rather than how fast it first paints.

CLS is not a mobile problem at all, by this data. Sites chasing a lower CLS score on mobile specifically are optimizing against the metric where mobile already outperforms desktop; the layout-shift fix, if one is needed, more often belongs on the desktop template.