llms.txt is a proposed convention: a plain-text (Markdown) file placed at a site's root that offers large language models a concise, curated map of the site's most useful content — a short description of what the site is, and a structured list of key pages with links and brief notes — so an AI system retrieving from the site can find the important, canonical material efficiently rather than crawling everything.
Definition
llms.txt is a proposed convention: a plain-text (Markdown) file placed at a site's root that offers large language models a concise, curated map of the site's most useful content — a short description of what the site is, and a structured list of key pages with links and brief notes — so an AI system retrieving from the site can find the important, canonical material efficiently rather than crawling everything. It is analogous in spirit to robots.txt or a sitemap, but aimed at LLM consumption and comprehension rather than search indexing.
It is an emerging, non-standard convention: adoption by AI systems is partial and evolving, it does not control training or access (robots.txt and server rules do that), and it is best understood as an optional aid for AI retrieval rather than a requirement or a ranking factor.
In context
For an iGaming affiliate, llms.txt is a low-cost, optional thing to add and worth doing if the site has a clear structure worth pointing an AI system to. A sensible file would describe what the site is (an iGaming information and comparison resource, with named editors and a methodology), and list the canonical hubs: the glossary A-Z index, key guides, the review methodology page, the responsible-gambling section, and the main market or topic hubs — each with a link and a one-line note.
The aim is that an AI retrieving from the site lands on the authoritative pages rather than thin or duplicate ones.
The cautions: it is not a substitute for AI crawler controls (if the site wants to limit training crawlers, that is robots.txt, not llms.txt), it is not guaranteed to be used, and it should not be treated as an SEO tactic or stuffed with commercial links. It should reflect the genuinely most useful, canonical content, be kept in sync as the site changes, and stay concise.
For a large glossary specifically, pointing to the A-Z hub and explaining the structure (bilingual, term pages with definitions, examples and related links, editor-attributed and dated) helps an AER system understand and cite it correctly. For affiliate-facing content, the framing is that llms.txt is an emerging, optional convention for guiding AI retrieval to a site's best content, that it does not control access or training, and that it is worth adding as a concise, honest map of the canonical hubs kept current.
Worked example
An affiliate adds an llms.txt at its root: a short description of the site as an editor-run iGaming information resource, then a linked list of the glossary A-Z hub, the review methodology, the responsible-gambling section, and the main market guides, each with a one-line note. It keeps the file updated and does not use it to push commercial pages.
Related terms
Frequently asked questions
Browse the full iGaming & affiliate glossary — hundreds of EN/RU terms with examples.
← Back to glossary