PageSpeed Insights

PageSpeed Insights

Team Pick
SEO & AEO
End the argument about whether a page is slow, using numbers from your own visitors.
ACCESS
No login
PRICE
Free
HOST
Google
CATEGORY
Page speed
Last updated
August 4, 2026

Preview

PageSpeed Insights

What it is.

Paste a URL and you get two answers: what Chrome users met over the previous 28-day collection period, and what a Lighthouse run says in a controlled environment you can repeat after every deploy.

Google publishes the thresholds, so you can check the verdict instead of trusting it. Largest Contentful Paint counts as good up to 2500ms, Cumulative Layout Shift up to 0.1, Interaction to Next Paint up to 200ms, each read at the 75th percentile of visits.

What you get.

  • A pass or fail on each Core Web Vital, measured on visits and not on a lab guess.
  • The share of visits that landed in the poor band, so you know who it affected.
  • The render-blocking files named, which is what makes the engineering ticket writable.
  • A rerunnable lab test, so you can show a change worked before field data catches up.
HOW TO USE IT

How to set it up.

1

Paste the URL and read the field data first, since the lab score is a simulation and the field data is what visitors met.

2

Switch between Mobile and Desktop, and treat the mobile numbers as the ones that count.

3

Compare each metric against the published thresholds: 2500ms for Largest Contentful Paint, 0.1 for Cumulative Layout Shift, 200ms for Interaction to Next Paint.

4

Open the diagnostics and work the render-blocking resources and image sizing items before anything else.

5

Rerun the lab test after each change, and wait for the 28-day window to catch up before you judge the field data.

Pricing Plans

Web interface

$0

no login, with separate mobile and desktop reports

Use cases

Settle the speed argument

Run the page, screenshot the field verdict, and put a published threshold next to the complaint.

Render-blocking list

Take the files delaying first paint to whoever owns the head of the page.

Before and after

Test the page before a change and again afterwards, so the improvement carries a number.

Best for

Settling a speed argument

Field data from Chrome visits ends the debate about whether a page is slow for users or only slow on someone's laptop.

Checking a landing page before launch

The lab run works on a page with no traffic yet, which is exactly when field data has nothing to show.

Prioritising engineering work

The diagnostics list names the files, so a ticket can carry the filename instead of the word slow.

Open the tool

Opens on Google in a new tab.

Open the Tool

Questions about PageSpeed Insights

Why do the two sections disagree?
Which number should I report?
Is there an API?

Questions about SEO & AEO

What counts as a good Core Web Vitals score?
Does page speed affect rankings?
Before you start
  • A publicly reachable URL, since it cannot open a staging site behind a login.
  • Enough live traffic for the field data to appear.
Strengths
  • Field data from Chrome visits over a trailing 28-day period, which no synthetic test can give you.
  • Published thresholds for every metric, so the verdict is checkable and not a vendor opinion.
  • No login, no quota on the web interface, and the same engine behind the Chrome DevTools audit.
Limitations
  • Field data needs enough Chrome traffic to appear, so a new or low-traffic page shows lab results only.
  • The 28-day window means a change you made last week is barely visible in the field numbers.
  • The lab score swings between runs on the same page, so read the metrics and not the headline number.
Skip this if
  • You already run speed monitoring on a schedule and get alerted when a page regresses.
Ideal for
  • SEO leads
  • Web teams
  • Anyone briefing developers

Got the free stack.
Need the engine?

Free tools give you the data. We turn it into pipeline for 300+ B2B tech companies, across organic search and paid media.