llms.txt: What It Is, and Should Your Website Have One?

What llms.txt actually is

llms.txt is a proposed web standard: a markdown file served at yourdomain.com/llms.txt that describes what your site is, what it offers, and links to your most important pages with short descriptions. Where robots.txt tells crawlers what they can't touch, llms.txt tells AI systems what they should read first — a curated front door instead of forcing models to reconstruct your site from raw HTML, navigation menus and cookie banners.

What it looks like

A minimal llms.txt is just structured markdown:

• An H1 with your site or business name
• A blockquote summary of what you do
• Sections linking key pages, each with a one-line description

You can see a live example on this very site: nakwave.services/llms.txt.

The honest part: does anything actually use it?

Here's what most agencies won't tell you: as of mid-2026, no major AI engine — not OpenAI, not Google, not Anthropic, not Perplexity — has officially confirmed that it consumes llms.txt for search or citation. Adoption is real on the publisher side (thousands of sites, including major developer-tool companies, ship one) but unconfirmed on the consumer side. Anyone selling llms.txt as a guaranteed visibility boost is overselling.

So why do we still ship it on every build?

• It costs almost nothing. Fifteen minutes of curation for a file that might matter — that's the cheapest option contract on the internet.
• The downside is zero. It can't hurt rankings, can't slow your site, can't confuse crawlers that ignore it.
• The curation exercise itself has value. Writing an llms.txt forces you to state clearly what your business is and which pages matter most — clarity that usually exposes gaps in the site itself.
• Adoption moves fast. If a major engine flips it on next quarter, sites that already ship it are grandfathered into the advantage; everyone else scrambles.

What matters more than llms.txt

If you're prioritizing, the levers with confirmed impact come first: complete JSON-LD structured data, answer-first content, real FAQ sections, fast Core Web Vitals and consistent entity information. llms.txt is the garnish, not the meal — we cover the full recipe in our AEO explainer and the practical steps in how to get cited by ChatGPT.