What is llms.txt (and does your site need one)?
robots.txt tells crawlers what they may read. llms.txt tells AI systems what your site actually is. It is a simple markdown file at the root of your domain, written for language models rather than browsers, and it takes about five minutes to add.
What it is and where it came from
llms.txt is a convention proposed in 2024: a markdown file served at yourdomain.com/llms.txt containing a short, factual summary of your site, its key pages, and the context a language model needs to describe you accurately. Think of it as a press kit for machines: who you are, what you do, who it is for, what it costs, where the important pages live.
The reasoning is practical. AI systems work under tight context limits and often cannot afford to crawl and digest your whole site. A clean, canonical summary in a predictable location gives them the right facts in one read, instead of leaving them to reconstruct your business from a navbar and a hero headline.
Honest answer: does it work?
Adoption is real and growing, and the major AI crawlers do fetch it. What nobody can promise is a measurable ranking effect, the ecosystem is too young and the engines do not publish their recipes. Our position is simple: it costs five minutes, it cannot hurt, it is read by the systems you want describing you correctly, and sites competing for AI citations should not skip free surface area. We ship one on drazen.ai ourselves.
- One-line description of what your product is, in plain language
- Key facts: who it is for, what it does, real pricing
- Links to your most important pages with one-line descriptions
- No marketing fluff: models quote facts, not adjectives
- Keep it current: an outdated llms.txt confidently feeds AI the wrong facts
Adding one
Write the file, serve it at /llms.txt as plain text, done. The subtle part is writing a summary that is genuinely quotable: specific, factual, and matching what the rest of your site says. Drazen generates and ships one as part of its fix set, keeps it consistent with your pages, and flags it if a rebuild deletes it. The free scan tells you in 30 seconds whether you have one and what else AI engines are missing about you.
Questions
Is llms.txt the same as robots.txt?
No. robots.txt controls crawler access (what may be read). llms.txt provides content (what your site is). They complement each other and live side by side at the domain root.
Will llms.txt get me cited by ChatGPT?
By itself, no. Citations come from being readable, clearly described and mentioned elsewhere on the web. llms.txt is one cheap piece of that puzzle: it makes your description accurate wherever the engine does read it.
What should I NOT put in llms.txt?
Marketing superlatives, keyword stuffing, or claims that contradict your site. Models cross-check; inconsistency reads as unreliability. State facts a careful journalist would be comfortable quoting.
See what your site scores
Drazen scans your site the way Google and ChatGPT see it, shows you everything that is hurting you, and writes the fixes itself. Free, about 30 seconds.
Run the free scan →