Free Performance SEO Tools

Lab scores and HTML checks are different jobs

PageSpeed and Core Web Vitals send the URL you paste to Google PageSpeed Insights. You get a Lighthouse lab run and, when Google has enough Chrome User Experience Report history for that URL, field numbers for LCP, INP, and CLS. Google's published Good thresholds are LCP under 2.5 seconds, INP under 200 milliseconds, and CLS under 0.1. If field data is missing, the page is new or low-traffic enough that Google has not published a CrUX sample. The tool should say so. It does not invent a field value.

The image and mobile tools never call PageSpeed. They read the HTML response: alt text, declared dimensions, file weight for images the checker can fetch, loading="lazy", viewport, and a few other markup signals. A hero with loading="lazy", or a large image without width and height, is a common reason the lab score and the HTML report disagree. If those HTML checks are clean and the lab score is still poor, the cause is usually script weight or a slow server response, which an alt report cannot see. Read the PageSpeed recommendations next.

Each run is one snapshot. It is not a 28-day field report and it is not a monitor. Lab scores move when the page, the network, or the CPU throttle changes. The mobile HTML check is not Google's retired Mobile-Friendly Test, and it does not load external stylesheets. Use a phone for tap targets and font size that are not in the raw HTML.

Do not run all six tools on every URL by habit. If the lab LCP is fine, an image-size pass is optional. If LCP is the failing metric, the image tools are the next click, on the same URL, before you rewrite copy. Alt text is an accessibility and image-search check. It does not move LCP unless the image was also the wrong file.