AI search does not read llms.txt (yet)

llms.txt was proposed as a map of your site for AI systems: one small file listing what matters, written for machines. The idea is reasonable. The measured reality, as of 2026, is that AI search does not use it.

The best data comes from Ahrefs, who checked server logs across 137,210 domains: 97% of published llms.txt files were never fetched by anyone real, and the fetches that did happen came overwhelmingly from SEO tools checking whether the file exists. Google has said it plainly — no AI system currently uses the file — and Google's own documentation asks for nothing special: AI features read the same pages as classic search.

There is one real exception: coding agents. Tools like Claude Code and Cursor, when pointed at a site, do fetch llms.txt and use it. If developers are part of your audience, the file has a genuine, narrow job.

So the honest guidance is both-ways: publish llms.txt — it costs minutes and serves the agents that do read it — but build your discoverability the boring way that measurably works: being in the classic search indexes with complete, server-rendered HTML. That is what this site does, and this page is itself part of the experiment.