LLMs.txt
LLMs.txt is a proposed file that gives AI models a curated map of your site in markdown. Learn what it contains, who supports it, and whether to add one.
TL;DR
LLMs.txt is a proposed standard: a markdown file at the root of your domain that gives AI models a curated map of your most useful content. It lists your key pages with short descriptions, in a format designed to be read directly rather than crawled and parsed. The proposal is young and adoption by engine operators is far from universal, which makes it a low-cost bet rather than a settled requirement.
Why LLMs.txt matters
The argument behind it is context efficiency. A model working from your rendered HTML spends its budget on navigation, scripts and boilerplate. A model given a curated markdown index spends it on your actual content.
- Curation: You choose which pages represent you, rather than leaving an engine to infer importance from your navigation.
- Clean format: Markdown carries structure without the markup overhead of a rendered page.
- Description control: Each entry gets a sentence you wrote, describing what the page covers.
- Cheap to try: A single static file, no infrastructure, and nothing breaks if engines ignore it.
How to create an LLMs.txt file
Place it at the root
The file belongs at yourdomain.com/llms.txt, served as plain text or markdown.
Open with a heading and a summary
Your company name as the heading, followed by a short paragraph describing what you do.
Group links under sections
Docs, guides, product pages, policies. Sections give a model a sense of what kind of content each link leads to.
Describe every link
A markdown link followed by one clear sentence. The description is the part doing the work.
Include only pages worth citing
Documentation, comparisons, pricing, methodology. Leave out login pages, thin category pages and anything you would not want quoted.
Consider a full version
An llms-full.txt containing the actual content of key pages, for models that will read further than an index.
LLMs.txt vs. robots.txt
LLMs.txt: Tells AI models what is worth reading and describes it. Inclusive, curated, and advisory in the sense that nothing enforces it.
Robots.txt: Tells crawlers what they may and may not fetch. Exclusive, mechanical, and long established with near-universal support.
They solve opposite problems. Robots.txt restricts access; LLMs.txt recommends attention. A site can and usually should have both.
Publishing the file is easy; knowing whether it changed anything is not. Findrix checks llms.txt coverage as part of a 31-point technical audit, then tracks how often seven AI engines actually cite you against named competitors. Every gap comes with the fix already written: technical, content and off-site. The audit is free, takes about a minute, and requires no signup.
What to include in LLMs.txt
- Documentation: The pages that explain how your product works, which get cited disproportionately often.
- Pricing: With current figures, since this is the fact engines most often get wrong.
- Comparison pages: Where you explain how you differ from named alternatives.
- Methodology or about: What your company is and how it does what it does.
- Key guides: Your strongest explanatory content, not your entire blog archive.
Should you actually add one
Yes, with realistic expectations. The file takes an hour, costs nothing to host, and carries no downside. If adoption grows, you are already positioned; if it does not, you have lost an afternoon.
What it will not do is make you visible. No engine has committed to weighting llms.txt heavily, and treating it as an AI visibility strategy rather than a small hygiene item leads to disappointment.
The useful side effect is the exercise. Choosing which twenty pages represent your company tends to surface how much of your site nobody should be citing.
Frequently asked questions
Do AI engines actually read llms.txt?
Support varies and no major engine has published a firm commitment to weighting it. Some tools and crawlers look for it. Treat it as a low-cost bet on an emerging convention rather than a mechanism with proven effect.
What is the difference between llms.txt and llms-full.txt?
llms.txt is an index: links with descriptions. llms-full.txt contains the actual content of those pages inline, so a model can read your material without following links. Publishing both is common.
Will llms.txt improve my AI visibility?
On its own, unlikely. Visibility depends mostly on third-party sources mentioning you and on your pages being crawlable and specific. Think of llms.txt as removing a small obstacle rather than creating an advantage.