Download websites that use Quicklink

About Quicklink

Quicklink belongs to a small class of front end libraries that exist only to make navigation feel faster. Instead of waiting for a visitor to click, the library watches which links are currently in the viewport and quietly fetches those pages while the browser is idle, so the next click resolves from cache. The technique is old, the discipline is not. Prefetching everything wastes other people's bandwidth, so Quicklink checks connection quality and data saver preferences before it spends anything. Published by Google under the Apache 2.0 license and in circulation since 2018, the library is written in JavaScript and small enough to drop into an existing page without a build pipeline.

A site carrying Quicklink is saying something specific about the people who built it. Nobody installs a prefetching library by accident, and it rarely arrives bundled with a theme or switched on by a hosting default. Somebody measured navigation timing, decided perceived speed was worth an extra dependency, and made the change. The resulting population is useful to anyone selling performance tooling, real user monitoring, edge delivery or front end consulting, because those sites have already admitted that page speed is a budget line rather than an afterthought.

Version data adds a second dimension. We count 11 distinct releases in circulation and most often resolve 2.3.0, and for a library that is usually pinned to a fixed version at the moment a page is written, the release on a site works as a rough timestamp of when the front end was last rebuilt. A stale pin next to a modern framework is a site that was tuned once and then left alone.

Detection here is inferential, and it undercounts. Quicklink is straightforward to see when it loads as its own script from a public delivery network or a predictable path. When a team pulls it through a bundler and the code ends up inlined with everything else in an application bundle, the fingerprint disappears and the site looks ordinary from outside. Read the count as a floor on deliberate, visible usage rather than a census of every project carrying the package in a dependency file.

Which is also why checking one domain answers a different question than downloading the list. A single lookup tells you whether one site prefetches. The export turns the question around: which sites, in which countries, on which platforms, and beside which other tooling. Cross referencing a performance library against ecommerce platforms, tag managers, hosting providers and framework choices is where usable segments come from, and none of that assembles itself from typing domains into a lookup box one at a time.