llms.txt

What it means

llms.txt is a text file at the root of a website that introduces the site to language models and AI assistants in a short, readable form and links to its most important content. It is a proposed convention, similar to robots.txt for search engines.

The file is written in Markdown: a heading with the name, a short description and lists of links to the main pages with one sentence about each. Some sites also offer an extended version with the full text of their content so the model does not have to assemble it from individual pages.

The point is practical. Web pages are full of navigation, scripts and advertising, and a language model finds it hard to pick out what matters. Clean text with a clear structure helps it answer accurately and cite the right source.

How it works in practice

Our website offers llms.txt files in all four languages, with contacts, services, prices and the whole glossary. An assistant answering in Russian or Ukrainian can therefore quote our text in that language instead of translating Slovak terms itself.

When preparing the file for clients, we recommend including only verified, current data: the exact company name, company ID, address, services and prices. If prices or terms change on the website, they must change in llms.txt too, or assistants will spread outdated information.

llms.txt does not replace conventional SEO, the XML sitemap or structured data. It is an addition that helps where people look for answers through AI assistants, and it takes little work to prepare.

What to watch out for

Because it is a new, non-binding convention, not every system reads the file. Relying on it alone makes no sense. Good, clear and accessible content on the website itself remains the foundation.

Inconsistency is a risk too. If llms.txt says something different from the website, such as another price, an assistant may quote the wrong version. The file is best generated automatically from the same data as the website rather than written by hand.