Field note
Does llms.txt actually do anything
It appears on a lot of sites and in a lot of advice. The honest position is that it is a proposal, not a standard, and nothing important currently depends on it.
More than five years in personal branding. Runs atharvsankpal.com and works with founders and businesses in the US on their digital presence. Previously produced AI video for brands.
Published 2026-08-14
Researched against primary sources, drafted with AI assistance, then reviewed and approved by the named author. How we write · About the author
llms.txt is a proposed convention: a plain text file at the root of a site giving language models a curated summary and a set of links. It is not an adopted web standard, no major answer engine requires it, and there is no published evidence that having one causes a system to cite you. It is also cheap, harmless and takes minutes, so the reasonable position is to add one and expect nothing from it. What actually determines whether you are cited is upstream of it entirely: whether crawlers can reach your pages, whether your writing contains complete factual statements a system can lift and attribute, and whether independent sources corroborate what you claim. Treat llms.txt as tidy housekeeping rather than as a lever.
- A proposal, not a standard. Nothing currently requires it.
- No evidence it causes citation. Claims otherwise are ahead of the facts.
- Cheap and harmless. Add one, expect nothing.
- The real levers are crawler access, quotable writing and corroboration.
What it is
A plain text file at the root of a domain, written in markdown, containing a short description of the site and a curated list of links with one-line explanations. The idea is to give a language model a clean, authoritative summary rather than making it infer one from navigation and marketing copy.
As an idea it is sensible. A concise, factual statement of what an organization is and where the substantive material lives is genuinely useful, and writing one forces a clarity that most sites lack.
What it is not
It is not a standard. It is a proposed convention, and adoption by the systems it targets is the part that has not happened. No major answer engine documents it as a requirement or a ranking input.
It is not an access control. robots.txt governs whether a crawler may fetch your pages. llms.txt does nothing of the sort, and a site that blocks retrieval crawlers while publishing a beautiful llms.txt has achieved nothing at all.
And it is not a shortcut. Most of the advice recommending it implies a causal link to citation that nobody has demonstrated. Be suspicious of that, including when we would benefit from you believing it.
Whether to bother
Yes, briefly, and then move on.
It costs half an hour. If adoption arrives you are ready, and if it does not you have lost nothing but the time. Writing one also produces a side benefit worth more than the file: being forced to state plainly what you do and which pages matter usually exposes that your homepage does neither.
What it should not do is displace the work that demonstrably matters. If you are choosing between writing an llms.txt and checking whether your host is blocking retrieval crawlers, check the crawlers.
Questions people actually ask
Is llms.txt an official standard?
No. It is a proposed convention. It has not been adopted as a web standard and no major answer engine documents it as a requirement or an input to how sources are selected.
Will llms.txt get me cited by ChatGPT?
There is no published evidence that it does. Citation depends on retrieval access, whether your pages contain quotable factual statements, and whether independent sources corroborate them.
Should I add one anyway?
Reasonably, yes. It takes very little time, does no harm, and writing it forces useful clarity about what your site is for. Just do not let it substitute for the things that are known to matter.
Is llms.txt the same as robots.txt?
No. robots.txt controls crawler access and is a long-established protocol. llms.txt is a proposed convention for offering a summary and does not control access to anything.
References
- The llms.txt proposal documentation.
- RFC 9309. Robots Exclusion Protocol. Useful for the contrast with access control.
- OpenAI. Overview of OpenAI crawlers — GPTBot, OAI-SearchBot and ChatGPT-User. Anthropic. Crawler documentation — ClaudeBot, Claude-User and Claude-SearchBot. Google Search Central. Crawlers and user agents, including the Google-Extended control.
How this was made: researched against primary sources, drafted with AI assistance, then reviewed and approved by the named author before publication. Our editorial standards.
Worth doing the things that are known to work first.
Check which crawlers you allow →Related reading
Want the levers that actually move it?
We will show you what answer engines currently return for your name and which of the real levers is missing.
Book a 30 minute call