What llms.txt is
A proposed convention: a Markdown file at the site root summarising the site and linking its most useful pages for language models. It is not a standard any provider is obliged to read, and it is not a ranking signal.
What we measured
Requests to our own files by crawlers that passed the provider's documented check, 25.07.2026 to 21.09.2026:
| File | Verified requests | | --- | --- | | robots.txt | 438 | | ai-sitemap.xml | 90 | | sitemap.xml | 73 | | llms.txt | 14 |
Of the 14, OpenAI accounted for 7, Google for 5, Apple and Perplexity for one each. The crawlers are reading the files search engines have used for decades, and our pages.
Where the effort pays
Publish llms.txt — ours is at /llms.txt, and it costs nothing to keep. But the pages it points to are what carry the facts: identity in the HTML, facts in sentences, one canonical address per product. Those are read whether or not any crawler opens the pointer. The glossary entry is at /glossary/llms-txt; the evidence rules at /evidence.