Lighthouse 13.6 Update Pushes PageSpeed Insights Scores Higher

PageSpeed Insights now runs on Lighthouse 13.6.0, released Oct. 27, and the new simulation will generally produce higher scores for the same pages — without a single change to the code on them.

Chrome's Web Performance Developer Advocate, Barry Pollard, confirmed the rollout on Bluesky. He said the update improves the simulated First Contentful Paint (FCP) and simulated Largest Contentful Paint (LCP), both of which will be more accurate than before and generally result in higher scores. Pollard added that websites heavily dependent on JavaScript are more likely to see an impact.

To back that point, Pollard quoted a developer who reported large score gains for single-page apps since the update, citing react.dev as an example. Pollard opened his post with "True!", then added:

"Note this is not SPA-specific (and unrelated to other SPA things we've been working on), but more likely to impact JS-heavy sites."

What changed in the simulation?

By default, Lighthouse loads a page at full speed and then estimates how it would load on a slower connection. The throttling documentation states this simulated throttling matches PageSpeed Insights' setup. The version 13.6.0 release notes list several changes to Lantern, the simulation engine behind those estimates, aimed at improving accuracy.

Two changes stand out:

• Lighthouse 13.5 used each request's fetch priority to decide whether it blocked rendering. In 13.6, Lantern uses Chrome's own render-blocking label for each request, falling back to priority only when that label is missing.
• When the LCP element is text that appears in the same frame as the page's first content, Lantern now estimates that LCP from the same requests and page tasks it uses for FCP, instead of from a set that can include every request made before LCP.

The second change addresses a real failure case. Developer Ivan Akulov demonstrated in a GitHub issue how the old simulation handled a test page built with Framer whose only above-the-fold content was text, so one element counted as both FCP and LCP. Lighthouse reported a much later LCP than FCP for that text. Preloading the page's JavaScript modules improved interactive time for the real page but made simulated LCP worse. Pollard thanked Akulov in a reply, saying he "kicked off some of these improvements." The Lighthouse team closed the issue as completed in the Lantern upgrade listed in the 13.6 release notes.

What does the ARD audit change?

Lighthouse 13.6 also updates the Agentic Resource Discovery (ARD) audit introduced in 13.5. The earlier version looked for an ARD manifest under the old ai-catalog names even though the spec had changed to ard.json — so a site using only the new name without an Agentmap line in robots.txt would show as Not Applicable.

Version 13.6 now checks for an Agentmap line, a rel="ard" link tag or Link header, and /.well-known/ard.json before falling back to ai-catalog names. If it finds the manifest only at an ai-catalog location, it notes that ARD consumers are not required to check there.

Lab scores versus field data

The performance score in PageSpeed Insights comes from Lighthouse lab data: a simulated load on a single device under fixed network conditions. Google's documentation warns that "having good lab data does not necessarily mean real-user experiences will also be good." Field data in the same report covers real visits over the last 28 days, as reported through the Chrome User Experience Report (CrUX). Each report lists the Lighthouse version used, next to the emulated device.

This distinction now matters for reporting. Scores from before and after the 13.6 update were not measured with the same simulation, so part of a higher lab score on a freshly optimized page may come from the new version, and charts tracking lab scores over time may show a jump at the switch. Before claiming an improvement, practitioners should confirm both reports name the same Lighthouse version and compare field data.

Lighthouse's release notes say the team expects 13.6 to ship in the DevTools of Chrome 159, currently in Chrome's desktop Dev channel and Canary. As of Nov. 2, Chrome's release schedule lists Chrome 159 for stable release on Dec. 8. An open issue on Lighthouse's GitHub continues to examine cases where Lantern's simulated LCP runs far above measured LCP, using large JavaScript and SPA sites as examples.