llms.txt: what it is and whether to use it

llms.txt is a plain Markdown file at the root of your site (/llms.txt) that tells AI tools what your site is and which pages matter most, with a link and one line for each. It was proposed in 2024 as a convention, not a standard. It helps coding agents and AI assistants read documentation. It is not a proven way to get cited in ChatGPT, Gemini or Google AI Overviews, because none of the major answer engines has said it uses the file to pick sources.

So the short answer: add one if you have docs, an API or a product that people use through AI agents. It takes about an hour and costs nothing. Do not expect it to move your AI visibility on its own.

What the file looks like

The proposed format is simple Markdown:

# Your Product

> One or two sentences on what the product is and who it is for.

Optional paragraph with context an AI tool needs, such as pricing model or key terms.

## Docs

- [Getting started](https://example.com/docs/start): install and first project
- [API reference](https://example.com/docs/api): every endpoint with examples

## Optional

- [Changelog](https://example.com/changelog)

The pieces:

• An H1 with the name. The only required part.
• A blockquote summary. What it is, in a sentence or two.
• Sections of links, each with a short note on what the page covers.
• An "Optional" section for pages a tool can skip when it is short on space.

Some sites also publish llms-full.txt, one file with the full text of their docs, so a tool can load everything in one request.

What it does

• Gives agents a starting map. When a coding agent or an assistant with a browsing tool is pointed at your site, a clean list of the right pages beats crawling navigation menus and cookie banners.
• Saves context. Markdown links with one line each are far shorter than HTML pages, so an agent can read more of what matters.
• Lets you choose what is canonical. You decide which pages describe the product, instead of a tool guessing from old blog posts.

What it does not do

• It is not a ranking signal. Google has said that its AI features rely on normal Search systems and that you do not need special files for them. OpenAI, Anthropic and Perplexity have not said their answer engines read llms.txt when choosing what to cite.
• It does not replace access. If robots.txt blocks OAI-SearchBot, GPTBot, PerplexityBot or Googlebot, a list of links will not help.
• It does not replace the pages. An AI answer cites the page that answers the question. If that page does not exist, linking to it is not possible.

You may see claims that a large share of sites now use llms.txt, or that it lifted someone's AI traffic. Treat those with care. Adoption counts vary with who did the crawl, and a traffic change after adding the file usually has other causes.

Should you add one?

zaDojo publishes its own llms.txt, because people set it up through AI agents and those agents need to find the setup steps.

How to write one in an hour

1. List the 10 to 30 pages you would send a new customer: what the product is, pricing, setup, the main docs, and comparisons.
2. Write one line for each, saying what the reader will find there. Plain facts, no slogans.
3. Group them under a few H2 headings, and put the nice-to-have pages under "Optional".
4. Save it as llms.txt at the root of your domain, served as plain text.
5. Open it in a browser and check every link works. Update it when you add or remove key pages.

Keep it honest. If the file says one thing and your pages say another, the tool reads the pages.

To skip the blank page, the free llms.txt generator builds a draft from your sitemap, page titles and descriptions. Edit it down to the pages that matter before you publish it.

What moves AI visibility instead

If your real goal is to be named in AI answers, these matter more than the file (the AEO guide turns them into a one-week plan):

• Pages that answer the questions buyers ask, with the answer in the first paragraph.
• Third-party pages (lists, reviews, comparisons) that the engines already cite for your topic.
• Crawler access for the bots behind each engine.
• Measuring it. Track the questions on a schedule and see who gets named and which pages get cited. The guide to tracking your brand in ChatGPT walks through it.

Questions people ask

Does Google use llms.txt? Google has said its AI Overviews and AI Mode work from its normal index and that no special AI file is needed. Normal SEO is what counts there; see how to rank in AI Overviews and AI Mode.

Is llms.txt the same as robots.txt? No. robots.txt tells crawlers what they may fetch. llms.txt suggests what is worth reading. It cannot allow or block anything.

Does it need to be in the root folder? The proposal puts it at /llms.txt. Some tools also try other paths, but the root is where they look first.

For the wider picture, read what GEO is and GEO vs SEO vs AEO. zaDojo is €10 a month with €10 of usage included, and new accounts get €0.10 of trial credits with no card.