LLMs.txt NOPE for most cited domains in AI answers

written by Gagan Ghotra

Published On

Last Updated

LLMs.txt as of now is something which most AI companies aren’t respecting but IETF (Internet Engineering Task Force) is working on development of AI Preferences protocol (which involves some quite interesting Technical SEO concepts) but right now all of most cited sites in AI answers aren’t using LLMs.txt file.

This graph from Semrush + Statista analysis been floating around on X today! And even earlier today in a meeting someone brought this up to me.

This chart shows Google’s Al Mode, Al overviews, ChatGPT and Perplexity. Based on 150 thousand citations from 5,000 randomly selected keywords from Semrush database which is just limited dataset.

And a brief overview of info in this is

  • Reddit is the clear leader with 40.1% of citations, far ahead of others.
  • Wikipedia ranks second at 26.3%.
  • YouTube (23.5%) and Google (23.3%) are close behind, forming a mid-tier with Wikipedia.
  • Yelp (21.0%) and Facebook (20.0%) also feature prominently, indicating substantial reliance on social and review platforms.
  • Amazon accounts for 18.7% of citations, reflecting product and commerce-related information use.
  • The remaining sources are smaller: Tripadvisor (12.5%), Mapbox (11.3%), and OpenStreetMap (11.3%).

BUT?

Domain/llms.txt status
reddit.comNot found
wikipedia.orgNot found
youtube.comNot found
google.comNot found
yelp.comNot found
facebook.comNot found
amazon.comNot found
tripadvisor.comNot found
mapbox.comNot found
openstreetmap.orgNot found

None of these sites is actually using llms.txt

And even for the top 100 most visited sites in Aug 2025 per Ahrefs dashboard! All of those don’t have llms.txt file at the root of their domains.

Categories:

Leave a Reply

Your email address will not be published. Required fields are marked *