Free Performance tool

Website Speed Test

We load your page as a mid-range phone on a slow 4G connection, measure how fast the main content appears and how much the page jumps, and show what makes it heavy. When Google has real-user data for your site, you see that too.

Free. Usually done in under a minute.

What this tool checks

  • A phone on slow 4G

    The page loads in Chromium as a Moto G Power class phone (412 × 823) with 150 ms latency, about 1.6 Mbps down and 750 Kbps up, and a 4× slower CPU. This is the profile Lighthouse uses for mobile tests.

  • Largest Contentful Paint

    When the largest image or block of text in the first screen finished rendering, and which element it was.

  • Cumulative Layout Shift

    How much visible content moved unexpectedly while the page loaded.

  • Total Blocking Time

    How long scripts kept the browser too busy to respond to a tap: the part of each long task over 50 ms, added up. It is the lab stand-in for Interaction to Next Paint.

  • What the page downloads

    Total bytes and requests, split into images, scripts, CSS and fonts, and the share that comes from third parties.

  • Heavy images and render-blocking files

    Images over 150 KB or far wider than the space they are shown in, and the CSS and script files that hold up the first render.

  • Real-user data from Chrome

    Google’s Chrome UX Report for your origin on phones at the 75th percentile: LCP, INP and CLS, when Google has enough traffic for the site.

What it can’t check

  • Give the same numbers on every run. Lab results vary with the network, server load and third-party scripts, so compare trends rather than single runs.
  • Match PageSpeed Insights exactly. Both use a similar phone profile but run on different machines and networks.
  • Measure INP in the lab. INP needs real interactions, so the lab test shows TBT and the field data shows INP when it exists.
  • Show field data for every site. The Chrome UX Report covers the last 28 days and only origins with enough Chrome traffic.
  • Test more than the one page you enter. The paid Performance Fix Report tests 3 key pages and ranks the causes with fixes.

Why it matters

A page that feels instant on office Wi-Fi and a fast laptop can take several seconds on a mid-range phone with a weak signal. That is the situation this test recreates, because it is where heavy images and scripts do the most damage.

Core Web Vitals are Google’s measures of that experience: loading (LCP), responsiveness (INP) and visual stability (CLS). Google uses them in ranking, but says a good page experience doesn’t outweigh relevant content. Their bigger effect is on people: a slow or jumpy page loses visitors before they read anything.

The causes are usually concrete: an uncompressed hero image of several megabytes, a chat widget that loads before the content, a web font that hides text, or a banner inserted above the page after it appears. The test names the LCP element and the heaviest files so you can see which of these applies.

How to read the results

Largest Contentful Paint (LCP)
Good under 2.5 s, poor over 4 s. If the LCP element is an image, its size and loading priority matter most; if it is text, look at web fonts and render-blocking CSS.
Cumulative Layout Shift (CLS)
Good under 0.1, poor over 0.25. A score without units that combines how much of the screen moved and how far.
Total Blocking Time (TBT)
Good under 200 ms, poor over 600 ms. A high TBT predicts slow responses to taps for real users.
Bytes and requests
There is no official limit; look at which type dominates. Images are usually the largest share, and a high third-party share points to apps, tags and widgets.
Heavy images
Each image over 150 KB, or downloaded far wider than it is shown, with its file size and dimensions.
Render-blocking files
CSS and synchronous scripts the browser must download before it can show anything.
Field data (75th percentile)
Real Chrome users on phones over the last 28 days. The 75th percentile means three in four page loads were at least this fast. INP is good under 200 ms and poor over 500 ms.
No field data
Google doesn’t have enough Chrome traffic for your origin. It says nothing about your speed; rely on the lab results.

How to fix the most common problems

  • A slow LCP image

    Serve the hero image at the size it is displayed, using srcset, in WebP or AVIF. Don’t lazy-load it, add fetchpriority="high", and put it in the HTML as an img element rather than loading it from a script or a CSS background.

  • Content that jumps while loading

    Give images, videos and embeds width and height attributes or a CSS aspect-ratio. Reserve space for banners, cookie notices and ads, or show them as overlays that don’t push content down. For web fonts, use font-display: optional or a fallback font sized to match.

  • High Total Blocking Time

    Remove apps, tags and plugins you no longer use first; their scripts often stay after you stop using them. Load chat, reviews and analytics with defer or after the page is interactive. On Shopify, check Online Store › Themes › Customize › App embeds.

  • Render-blocking CSS and scripts

    Add defer to scripts that don’t need to run before the first paint, inline the small amount of CSS the first screen needs, and drop stylesheets for features the page doesn’t use.

  • Images that are too heavy

    Resize images before uploading (around 2000 px wide is plenty for a full-width banner) and compress them. Shopify and WordPress can serve smaller versions automatically when the theme asks for them, so check that your theme uses responsive image sizes.

Questions about this tool

Why are my results different from PageSpeed Insights?

Lab tests depend on the machine, the network and the moment. PageSpeed Insights runs Lighthouse on Google’s servers and simulates the slow connection, while we apply the throttling in a real browser. Expect similar ranges, not identical numbers.

What is a good LCP?

2.5 seconds or less for at least 75% of page loads, according to web.dev. Over 4 seconds is poor.

Why is there no real-user data for my site?

The Chrome UX Report only includes origins that are publicly discoverable and get enough visits from Chrome users who share usage statistics. Many small sites don’t reach that level; the lab test still shows what to fix.

Why test on a slow phone instead of a desktop?

Because problems show up there first. A fast laptop hides heavy images and scripts that cost seconds on a mid-range phone, and Lighthouse uses the same phone profile for its mobile tests.

What happened to First Input Delay?

Interaction to Next Paint (INP) replaced First Input Delay as a Core Web Vital in March 2024. INP looks at the delay of interactions throughout a visit, not only the first one.

Sources

Related tools

The Performance Fix Report ($55, one-time) starts from the full free scan and adds more pages, the change to make for every finding, a PDF and re-tests to verify your fixes.

About the Performance Fix Report