What is llms.txt — and should your site have one?
What is llms.txt?
llms.txt is a single Markdown file served at /llms.txt that points large language models to your most useful pages, with short descriptions, in a clean, link-first format. The goal is simple: web pages are cluttered with navigation, scripts and ads, so llms.txt offers AI engines a distilled, machine-friendly summary of what your site is and where the substance lives.
It was proposed in 2024 and has been adopted by a growing list of developer-focused sites. Importantly, it is a community convention, not an official mandate — adoption by AI engines is still uneven, so treat it as a complement to (not a replacement for) the fundamentals.
How llms.txt differs from robots.txt and sitemap.xml
| File | Audience | Job |
|---|---|---|
| robots.txt | All crawlers | Permission — what may be crawled |
| sitemap.xml | Search engines | Inventory — every URL to index |
| llms.txt | AI / LLMs | Curation — the best content, summarised |
What to put in your llms.txt
- 1An H1 with your brand name, then a one-line blockquote summary of what you do.
- 2A short paragraph of context — who you serve and what makes you credible.
- 3Curated link sections (e.g. Key pages, Services, Contact) with a sentence describing each link.
- 4Only your highest-value, evergreen URLs — this is a curated map, not a full sitemap.
- 5Plain facts an engine can quote verbatim: markets served, contact details, core offerings.
Should your site have one? An honest answer
For most brands serious about AI visibility, yes — because the cost is near zero and the downside is none. It will not, on its own, make ChatGPT cite you; citations still come from clear on-page structure, schema, crawler access and entity authority. Think of llms.txt as one tidy signal in a larger GEO programme, not a magic switch.
“llms.txt won’t get you cited by itself — but a site with clean structure, schema and an llms.txt is unmistakably easier for an AI engine to trust.”
Frequently asked questions
No. llms.txt is a community-proposed convention introduced in 2024, not an official standard ratified by search engines or a standards body. Some AI tools and developer platforms support it, but adoption is still emerging and uneven. It is safe and cheap to add, but you should not rely on it as your only AI-visibility measure.
Not on its own. Being cited inside AI answers depends mainly on clear page structure, accurate schema, crawler access (robots.txt allowing AI bots) and your entity authority across the web. An llms.txt can make your best content easier for engines to find and summarise, so it complements those signals — but it is not a guarantee of citation.
Place it at your site root so it is reachable at https://yourdomain.com/llms.txt, the same location pattern as robots.txt. It should be a plain Markdown file. Many sites also publish a fuller llms-full.txt with expanded content, while keeping llms.txt as the concise, curated index.
The AI Search team at Caught by Apricot helps brands become the cited answer inside ChatGPT, Gemini, Perplexity and Google AI Overviews through structure, schema and entity authority.