Website Speed Test

Enter a page URL above, choose a location and device, then run the test. Calibre loads the page in Chrome and returns load timings, Lighthouse audits, a render video and a log of every request. It is free, needs no sign-up, and each report stays shareable for 30 days.

One page load in Chrome under fixed conditions, with Time to First Byte, First Contentful Paint, Largest Contentful Paint, Total Blocking Time and Cumulative Layout Shift. You also get Lighthouse audits for Performance, Accessibility, SEO and Best Practices, and a ranking of the third party scripts that slow the page most. Where Google has enough data, the report adds the Core Web Vitals real Chrome visitors recorded over the previous 28 days.

This test uses the same desktop and mobile settings and simulated throttling as PageSpeed Insights, so results are comparable. The main difference is location: PageSpeed Insights picks it for you, while here you choose Sydney, Frankfurt or North Virginia. Add your site to Calibre for scheduled tests from 17 locations, Chrome User Experience Report data for its important pages, and Real User Monitoring.

Use Google’s thresholds rather than one load time: Largest Contentful Paint of 2.5 seconds or less, Cumulative Layout Shift of 0.1 or less, and Total Blocking Time under 200 milliseconds, the lab stand-in for Interaction to Next Paint. Aim for Time to First Byte of 800 milliseconds or less and First Contentful Paint of 1.8 seconds or less. A Lighthouse Performance score of 90 or above is good.

The mobile test emulates a Moto G Power phone on a Slow 4G connection, while desktop uses an unthrottled processor and a much faster connection. On a slower processor and network, JavaScript and large images cost more. A large gap between the two often points to heavy JavaScript.

No two page loads are identical, because server response, third party scripts and advertising vary, so small differences between runs are normal. Larger differences between tools come from location, emulated device and network, throttling method and Lighthouse version. Run the test a few times and compare the pattern.

Start with the metric furthest from its threshold. Then check the Lighthouse audits with the largest estimated savings and the third party scripts that cost the most. Change one thing, retest from the same location and device, and compare.