// Tool · Performance · Free
Core Web Vitals check.
Enter a URL and read its LCP, CLS and INP on mobile. The numbers come from Google PageSpeed Insights, using real Chrome user data where Google has enough of it and lab results where it does not.
// Your page
// Result
Enter a URL and run the check to see LCP, CLS and INP.
Powered by Google PageSpeed Insights. Delivix stores nothing from your request.
// How it works
Where the numbers come from.
Your browser calls the public PageSpeed Insights v5 endpoint directly. Nothing passes through a Delivix server, and no API key is attached, so the request draws on a free allowance shared with everyone else calling it the same way. Busy days use that up, and the tool tells you when that has happened.
Field data first
If Google has enough real Chrome traffic for the URL, it returns the 75th percentile of what actual visitors experienced over a rolling 28-day window. That is the data Google uses in search, and it is what this tool shows whenever it exists.
Lab data as the fallback
Low-traffic pages have no field data. In that case the tool reads the Lighthouse run instead: one simulated mobile load on a throttled connection. It labels the result as lab data and substitutes Total Blocking Time for INP, because INP cannot be measured without real interactions.
The thresholds
- LCP: good at 2.5 seconds or under, needs work up to 4 seconds, poor beyond that.
- CLS: good at 0.1 or under, needs work up to 0.25, poor beyond that.
- INP: good at 200ms or under, needs work up to 500ms, poor beyond that.
- Total Blocking Time, the lab stand-in: good at 200ms or under, needs work up to 600ms.
What to do with a poor result
A single number tells you there is a problem, not what it is. The full PageSpeed report lists the failing audits, and the fix is usually images, fonts, third-party scripts or main-thread JavaScript. Working through that properly is our website speed optimisation service, and the same work is built into every site in our web design projects from the start.
Web Vitals, answered.
Largest Contentful Paint measures how long the main content takes to appear, and Google wants 2.5 seconds or less. Cumulative Layout Shift measures how much the page jumps around while loading, with a target of 0.1 or lower. Interaction to Next Paint measures how quickly the page responds when someone taps or clicks, with a target of 200 milliseconds or less.
Field data comes from real Chrome users who visited your page over the last 28 days, and it is what Google uses in search. Lab data is one simulated mobile load on Google hardware, useful for diagnosing a cause rather than measuring the experience. Pages with low traffic have no field data at all, and this tool falls back to lab results and tells you when it does.
Interaction to Next Paint only exists in field data, because it needs real people interacting with the page. When a URL has too little traffic for field data, the tool shows Total Blocking Time from the lab run instead. Total Blocking Time is a proxy rather than the same metric, so treat it as a signal about main-thread work, not an INP score.
The PageSpeed Insights endpoint is free and keyless here, so Google rate limits it against an allowance shared by everyone using it that way. When the allowance is spent the request comes back rejected, and the tool asks you to wait a minute and try again. It also fails on pages behind a login, pages blocked by robots.txt, and anything not publicly reachable.
They are a confirmed but light ranking signal, and they matter most as a tie-breaker between pages of similar relevance. The larger effect is on conversion. A page that takes four seconds to show its content loses visitors before the ranking has a chance to pay off.
In our experience the usual culprits are oversized images, render-blocking third-party scripts, fonts that load late, and layout that reserves no space for images or ads. LCP responds to image and font work, CLS to explicit width and height attributes, and INP to cutting JavaScript that runs on the main thread. All of it is measurable before and after.
// Next step
Knowing the score is step one.
We take failing pages apart, fix what is actually slow, and show you the before and after.