llms.txt: does it work? What it is, what the evidence says, and the file I use

If you read about AI search, you will meet llms.txt quickly. Some guides call it essential; others say it does nothing. The honest answer sits between the two.

What llms.txt is

llms.txt is a plain Markdown file you place at the root of your site, at yourdomain.com/llms.txt. It is a proposal, not a standard that every AI company has agreed to follow. The idea is to give language models a short, curated summary of who you are and which pages matter most, in a format that is easier to read than a full website with menus and scripts.

It is different from robots.txt. A robots file tells crawlers what they may access. llms.txt does not block or allow anything; it only suggests.

Does it work?

The honest answer is that there is little evidence it improves visibility in AI answers. Guides such as Core dna's "do you need it" and Static's guide take a cautious line, and several sources report that major AI vendors have not confirmed they use the file. One widely repeated claim from a large-site study is that most llms.txt files received no requests from AI bots over a month, and that some of the requests that did arrive came from SEO tools checking the file.

I cannot verify those logs myself, so treat that as a reported finding, not a fact. What I can say is that nobody has shown a reliable ranking or citation benefit, and anyone promising one is guessing.

What Google itself says

Google's AI features documentation states that there are no additional requirements to appear in AI Overviews or AI Mode, and that you do not need to create new machine-readable files, AI text files or markup for them. It points back to standard SEO fundamentals: crawlable, indexed pages with useful text content. So llms.txt is not something Google asks for. Whether other AI tools read it is a separate, unproven question.

Who should consider adding one

• Developer and documentation sites. Coding assistants and agents may read a clean index of docs, and this is where support is strongest.
• Small sites that want a tidy summary. It takes a few minutes and costs nothing.
• Anyone who can keep it up to date. A stale file can mislead.

Who should not prioritise it: a local business with thin content and no technical SEO. Fix crawlable pages, titles and structured data first.

The file I use

My site has one at gauravsah.com.np/llms.txt. I added it because it costs almost nothing, and I am not claiming it has changed any ranking or citation. The structure is simple:

1. A title. My name as a top-level heading.
2. A one-paragraph summary in a quote block: who I am, where I am based and what I do.
3. Contact and profile links.
4. A "Pages" section listing the main pages with one line each.
5. A "Case studies" section with factual one-line descriptions of the work.

Every line is something I can stand behind. No keyword stuffing, no claims I cannot support.

How to write a good one

• Write for a reader who knows nothing about you.
• Link only to pages you want found and that return a normal 200 response.
• Keep descriptions factual and short.
• Update it when your pages change.
• Do not copy a competitor's file or fill it with keywords.

What to do first, and what to measure

llms.txt should sit behind the fundamentals:

• crawlable, fast pages with one clear heading each,
• accurate titles and descriptions,
• structured data such as Person or Organization,
• an XML sitemap and a robots.txt that does not block the crawlers you want,
• clear, specific content that answers real questions.

If you add the file, check your server logs for requests to /llms.txt and see which user agents ask for it. That tells you more than any opinion piece, including this one.

For a wider view of getting cited in AI answers, read how to get your startup cited by ChatGPT and Perplexity. If you want an AI-search-ready setup for a small team, see my services.