Core Web Vitals checker: interpret your LCP, INP and CLS

How to use this tool

1. Open the report that contains your measurements and check its device category, collection period and URL or origin scope.
2. Enter LCP in seconds, INP in milliseconds and CLS as a unitless score. Do not substitute a Lighthouse performance score for a metric value.
3. Select whether the measurements come from field data, a lab test or a source you have not yet identified.
4. Run the checker and review each rating. Keep the original report so you can return to its diagnostic details.

You do not need to invent missing values to use the checker. Leave an unavailable metric blank. A missing INP value is not zero, and a Total Blocking Time value is not a replacement for INP.

About the results

The checker compares the values you enter with Google's published thresholds. Good values are LCP at or below 2.5 seconds, INP at or below 200 milliseconds and CLS at or below 0.1. Values above 4.0 seconds, 500 milliseconds and 0.25 respectively are poor; the intervening ranges need improvement.

The tool does not visit a URL or collect new performance measurements. Its calculation runs in your browser. It cannot verify that a pasted value represents the correct percentile, device group or page. A good rating means the entered number is in the good range, not that this tool has independently certified the website.

For a full field assessment, compare the relevant 75th-percentile values. The threshold reference explains the boundaries, while the beginner guide explains how to obtain and interpret measurements.

Why can this rating differ from PageSpeed Insights?

First check the inputs. PageSpeed Insights may show real-user values for an origin when a specific URL lacks sufficient data. Its lab section describes another measurement entirely. Mobile and desktop results can also differ.

Missing-data rules matter too. PageSpeed Insights can assess LCP and CLS when INP data is insufficient, whereas this checker can only classify the fields you supply. Preserve the source report's assessment and coverage notes rather than assuming the two displays must have identical wording.

Can I enter values from a lab recording?

Yes, provided you treat the result as a rating of that test. A controlled interactive browser recording can measure interaction latency. A standard Lighthouse page-load run does not measure field INP and instead uses TBT as a lab indicator of blocking.

One test is not a p75 population result. Use lab evidence to reproduce a problem, then confirm deployment outcomes with field data where available. The field versus lab guide explains why those results can disagree.

Does a good rating guarantee higher rankings?

No. Core Web Vitals are one part of Google's broader assessment of page experience, and good report results do not guarantee search rankings. Use the rating to identify an experience problem and choose a diagnostic path. Start with the failing metric and representative pages rather than making changes solely to improve a color.