Download websites that use Gutenberg

About Gutenberg

Versions are where Gutenberg becomes interesting, because most of the technologies we track report one number or none at all. Here the data shows 279 distinct builds in circulation, with 20.7.0 recurring across the population. A spread that wide on a single editor is not disorder for its own sake. The editor ships bundled inside the content management system and also exists as a separate plugin that runs ahead of the bundled copy, so every site carries whichever combination its owner chose or stopped choosing.

Version distribution is a behavioural measurement wearing technical clothing. A site on a recent build updates, which means somebody maintains it, which means there is a budget and a person attached. A site years behind has been left alone, and a site left alone accumulates other things: outdated extensions, expired integrations, and eventually an incident. The version column converts a list of domains into a list of maintenance postures, and maintenance posture predicts most of the commercial questions worth asking about a website.

Honesty about the measurement matters here. Build numbers are read from asset paths and markup that a site can strip, cache or rewrite, and optimisation tools that merge and rename files remove the evidence altogether. Sites with aggressive performance tuning are less likely to report a build at all, which biases the version split toward less tuned installations. The count of sites is more dependable than the split between builds, and any analysis resting on the split should carry that warning alongside it. Direction still holds even where the level does not, so a build losing share across releases is a safer conclusion than the share itself.

As of September 2026 we detect the editor on 909,784 sites, with 372,245 of them in United States and .com the recurring suffix. Automattic drives development with a large outside contributor base and releases under GPL v2, which is exactly why the plugin and the bundled copy diverge. Development moves at a pace core releases cannot match, so an owner who wants newer blocks installs the standalone plugin and inherits a faster update cadence than the rest of the site runs on. A domain doing that is telling you something about its owner's appetite for change.

The question the download answers is not whether a given site uses the block editor. A lookup settles that in a second, and the answer is usually yes for anything built on the platform in recent years. The real question is distributional. How many sites in a country are stranded on old builds. Which builds coexist with which commerce extensions. Which cohort updated within weeks of a release and which never moved. Whether the population running the standalone plugin behaves differently from the population on the bundled editor, which is a question about upgrade risk that agencies, hosts, extension vendors and security teams all pay money to answer. Distribution needs the whole set rather than a sample of it, and the whole set is what the export contains. Sampling a population this fragmented reproduces the fragmentation badly, which is the practical reason the file exists in the first place.