Google Lighthouse update changes how PageSpeed Insights simulates loading

PageSpeed Insights now runs Lighthouse 13.6.0, released on October 27, 2026. The new version changes Lantern, the simulation that estimates how long a page takes to load on a slower connection. According to Barry Pollard of the Chrome team, simulated FCP and LCP are now more accurate and scores are usually higher. A page's PageSpeed Insights score may therefore go up even though nothing on the page has changed. The change mainly affects lab data, not data from real users.

What does the Lighthouse 13.6.0 update change?

Lighthouse 13.6.0 mainly improves Lantern, the engine that simulates page loading. According to the release notes, the goal of the changes is better simulation accuracy.

By default, Lighthouse loads a page at full speed and then estimates how it would load on a slower connection. This mode is called simulated throttling. The Lighthouse documentation says it matches the PageSpeed Insights setup.

Previously, Lighthouse 13.5 judged whether a resource blocks rendering based on its fetch priority. Version 13.6 uses the flag set by the Chrome browser itself. Priority now serves only as a fallback when that flag is missing.

One of the changes was started by Ivan Akulov in a GitHub issue. He showed a page built in Framer where the same text was both the FCP and the LCP element. The old simulation reported a much later LCP than FCP for it. The Lighthouse team closed the issue as resolved in release 13.6.

How does the change affect PageSpeed Insights scores?

The change can raise a page's performance score in PageSpeed Insights without any changes to the page itself. According to Pollard, simulated FCP and LCP are now more accurate and usually produce higher scores.

Pollard noted that the change does not apply only to single-page applications (SPAs). JavaScript-heavy sites are more likely to notice it, though. He said this in reply to a developer who saw a big improvement in SPA scores, including the react.dev site.

The release notes said version 13.6 would reach PageSpeed Insights within a week. It is due to reach Chrome DevTools with Chrome 159.

Lighthouse 13.6 also changes the Agentic Resource Discovery (ARD) audit. It was aligned with the ARD v0.91 specification, in which the ard.json file replaced the ai-catalog.json name. The audit now looks for the manifest in places such as the HTTP Link header and at /.well-known/ard.json.

What does this mean for SEO and Core Web Vitals?

The update changes lab data, but not data from real users. This matters when you interpret the results.

PageSpeed Insights shows two kinds of data. According to Google's documentation, lab data is a simulated load on a single device under fixed network conditions. Field data comes from real users, on different devices and connections.

The Core Web Vitals report in Google Search Console is based on data from real users. A higher lab score after the update therefore does not automatically mean better Core Web Vitals. Google notes in the PageSpeed Insights documentation that values from the two sources can differ.

What should you do after your PageSpeed Insights scores change?

When your scores change, check whether the reports you are comparing were produced with the same Lighthouse version. Scores from before and after the update don't come from the same simulation.

Every PageSpeed Insights report shows the Lighthouse version next to the emulated device information. This makes it easy to check which simulation a score comes from. PageSpeed Insights scores also often go into SEO audits, so when you compare successive audits, it's worth recording the Lighthouse version.

The Lighthouse team is still working on the simulation. There is an open GitHub issue about cases where simulated LCP is much higher than measured LCP. The examples given are large JavaScript-based sites and SPAs.

We read PageSpeed Insights scores together with Search Console data in our SEO services. If you run a store, the free store UX audit walks through its mobile purchase path. What an SEO budget depends on is covered in our article on how much SEO costs.

What sources is this article based on?

This article is based on Matt G. Southern's piece “Google Lighthouse Update Changes How PageSpeed Insights Simulates Loading” in Search Engine Journal, published on November 2, 2026. The Lighthouse changes are confirmed by the 13.6.0 release notes in the GoogleChrome/lighthouse repository on GitHub. The differences between lab and field data are described in Google's “About PageSpeed Insights” documentation. Barry Pollard's statements come from his Bluesky posts, quoted by Search Engine Journal.