Largest Contentful Paint
When the largest visible content element finishes rendering. A field-data good threshold is 2.5 seconds or less at the 75th percentile. One Lighthouse run does not determine field status.
Free public-page analysis
Turn a speed score into a next step.
Test a public URL with Google PageSpeed Insights. We organize the Lighthouse result into a readable snapshot and highlight where to investigate first.
Run a new test
Enter a public page
Use a representative page, not only the homepage. Product, checkout, article, and landing pages often behave differently.
Protect private information. The submitted URL is
sent directly from your browser to Google PageSpeed Insights. Test
only a public URL you own, operate, or are authorized to test. Do
not enter internal, password-protected, or confidential URLs, or
links containing access tokens, API keys, personal information, or
other secrets.
Testing your public page
Connecting to the test engine…
A full Lighthouse run can take up to a minute. Keep this tab open.
Your performance snapshot
Open full report ↗
Results for
·
Measured lab signals
Highest-impact performance opportunities
Results are estimates under simulated conditions. They can vary by server load, page content, cache state, and third-party services. Performance data is provided by Google PageSpeed Insights.
Metric field guide
What the lab signals mean.
Use the numbers to locate a class of problem. Then verify the cause in the detailed report before changing infrastructure.
Total Blocking Time
Lab-only evidence of main-thread work that can delay interaction. Heavy JavaScript is a common cause.
Cumulative Layout Shift
Unexpected movement while the page loads. A field-data good threshold is 0.1 or less at the 75th percentile. One lab run is a diagnostic snapshot.
First Contentful Paint
When the first text, image, or other visible content appears. It helps explain how quickly the page starts to feel active.
Speed Index
How quickly visible content is populated during the lab run. Lower is generally better.
Server response
A slow initial document response can point to origin, application, database, geographic, or cache issues—but it is not proof of one.
Lab vs. field
One page has two performance stories.
Lab data is controlled and diagnostic. Field data describes experiences collected from real visitors over time. They answer different questions.
Lab data
A simulated load in a defined environment. Useful for debugging, repeatable comparison, and identifying technical opportunities.
Field data
Aggregated real-user experience when enough eligible traffic exists. Useful for understanding how the site performs across actual devices, networks, and locations.
Hosting or page problem?
Open the hosting guide
Separate infrastructure from page-level work.
Use our hosting guide to map a slow response, traffic constraint, or operational burden to the right hosting model.