Download websites that use Twitter

About Twitter

The name recorded here and the name on the product no longer match, and the gap is a good place to start. Twitter is owned by X Corp and has been rebranded, while the code embedded in millions of pages still carries the original naming in its script paths, its markup and its widget attributes. Detection follows what pages actually load, so the older label stays accurate as a description of the artefact even after the company moved on.

Understanding what that artefact is keeps expectations sensible. Launched in 2006 and offered as a Proprietary platform, the service reaches websites through a small JavaScript payload rather than through any deep integration, which means a share button, a follow button or an embedded timeline. Adding one takes a minute. Removing one takes a deliberate act that nobody is ever paid to perform, and that asymmetry is the single most important thing to know before using the data.

As of September 2026 we detect the embed on 189,702 sites, with 83,648 attributed to United States and .com leading the domain endings. Presence at that scale is not evidence that each site runs an active social programme. A great many of the markers are ordinary sharing furniture placed in a theme years ago and inherited by every page since, which is why the population should be read as sites that once decided the platform was worth linking to, and never as a list of accounts posting this week.

Used with that caveat in place, the signal is genuinely useful. Social embeds are one of the clearest ways to observe how a consumer platform propagates into the wider web, and the population moves in response to real events such as ownership changes, pricing changes and API restrictions. Comparing snapshots over time shows removal as well as adoption, and removal is often the more interesting direction. Marketing technology vendors use the presence of embeds to size the market for their own sharing and syndication products. Researchers use it to measure platform reach outside the platform. Performance and privacy teams use it because third party social scripts sit squarely in the middle of both conversations.

Detection has limits worth stating. A page that mentions an account in plain text, or links to a profile without loading the embed code, produces no marker at all, so a site with a very active presence can be absent from the list. Tag managers and consent tools that defer or block third party scripts can also hide the payload from a crawler, which pushes the count downwards. What we report is the presence of the embed on a public page, not the existence of an account behind it.

A single lookup tells you whether one site carries the widget, which a browser can tell you anyway. The downloadable population answers the questions that need the whole set, such as which sectors and countries still carry social embeds, what analytics and consent stacks travel alongside them, and which sites dropped the code between one release and the next.