llms.txt is a proposed standard for a Markdown file at the root of your site, at /llms.txt, that gives large language models a clean, curated map of your most important content. The idea, proposed in 2024, is to help AI systems find and understand your key pages without wading through navigation, scripts, and clutter. It is a proposal, not an official requirement from any AI provider.

Does llms.txt actually work in 2026?

The honest answer is that adoption is unconfirmed. As of 2026, major AI assistants have not publicly committed to consuming llms.txt, and there is no reliable public proof that adding one changes how often you are cited. It is low cost and low risk to add, but treat any strong claim about its impact with skepticism.

llms.txt: What It Is, Whether It Works, and How to Add It

Q: Is llms.txt the same as robots.txt?

No. robots.txt tells crawlers which paths they may or may not access. llms.txt is a curated content map meant to help language models find and understand your important pages. robots.txt is widely supported by compliant crawlers, though it is voluntary and not a security control. llms.txt is a newer proposal with uncertain adoption.

Q: Should I add an llms.txt file?

If it costs you an hour, adding a clean llms.txt is a reasonable low-risk bet, especially for documentation-heavy sites. Just do not expect it to be the thing that wins you AI citations. Prioritize extractable content, consistent brand facts, and third-party mentions first, and treat llms.txt as a small optional extra.

llms.txt is a proposed standard for a Markdown file at the root of your website, served at /llms.txt, that gives large language models a clean, curated list of your most important pages. The goal is to help AI systems find and understand your key content without parsing navigation, scripts, and clutter. It was proposed in 2024 and has spread in developer and documentation circles since. The honest caveat up front: it is a proposal, not an official requirement, and its real-world impact in 2026 is unconfirmed.

I’m Andrii Byzov, a fractional CMO for B2B tech. Teams keep asking whether they should add one, so here is a straight answer: what llms.txt is, the real state of adoption, how to add it, and how much to care. For the broader picture of optimizing for AI answers, see getting recommended by ChatGPT, Perplexity, and AI Overviews.

Key takeaways

llms.txt is a proposed Markdown file at /llms.txt that maps your key pages for language models.
It is not robots.txt. One is a curated content map, the other is crawler access control.
Adoption is unconfirmed. No major assistant has publicly committed to consuming it, and proof of impact is thin.
It is low cost and low risk to add, which is the main reason to bother.
Do the high-impact GEO work first. Treat llms.txt as a small optional extra, not a strategy.

What llms.txt actually is

The proposal is simple. You publish a Markdown file at the root of your domain that lists your most important URLs with short descriptions, optionally grouped into sections, sometimes with a companion file that inlines the full text of key pages. Markdown is used because it is clean and easy for models to parse. Think of it as a hand-curated table of contents written for machines rather than a sitemap, which lists everything, or robots.txt, which controls access.

A minimal example looks like this:

# Your Company

> One sentence on what you do and who you serve.

## Core pages
- [What we do](https://example.com/services): Short description.
- [Pricing](https://example.com/pricing): Short description.

## Documentation
- [Getting started](https://example.com/docs/start): Short description.

The honest state of adoption in 2026

This is where you should be skeptical of confident claims. As of 2026, the major AI assistants have not publicly confirmed that they read llms.txt to build answers, and there is no reliable public evidence that adding one increases how often you are cited. Some documentation platforms and developer tools now appear to generate llms.txt automatically, so the format has momentum in that world, but momentum is not the same as confirmed consumption by the assistants your buyers use.

So treat llms.txt as a reasonable, cheap experiment, not a proven lever. If a vendor or post tells you it will get you cited in ChatGPT, ask for the evidence. The defensible position is that it is unlikely to hurt if accurate and maintained, it might help machines navigate your content, and the impact on AI citations is unproven.

How to add one

List your most important pages. Not everything. The pages you would want an AI to understand first: what you do, pricing, core docs, flagship articles.
Write it in Markdown. A title, a one-line summary, then grouped links with short descriptions, as in the example above.
Serve it at the root. Publish the file so it resolves at https://yourdomain.com/llms.txt.
Keep it current. A stale map can be worse than none. Update it when your key pages change.
Do not duplicate robots.txt or your sitemap. This is a curated highlight reel, not a full index or an access-control file.

For static sites this is often a one-file change. For larger sites, some platforms and plugins can generate it from your existing content.

Where it sits in your priorities

If adding a clean llms.txt costs you an hour, it is a fine low-risk bet, especially for documentation-heavy products. Just keep the order of operations honest. The work most likely to improve AI visibility is extractable content, consistent brand facts, and third-party mentions, which I cover in LLMO vs GEO vs SEO and in my AI search visibility audit. Do those first. Add llms.txt as a small extra, measure what you can, and do not let a tidy text file stand in for the harder, higher-impact work.

If you want help prioritizing the AI-visibility work that actually moves the needle, you can reach me on LinkedIn.

FAQ

What is llms.txt? A proposed Markdown file at /llms.txt that gives language models a curated map of your most important pages. It is a proposal, not an official requirement.

Does llms.txt work in 2026? Adoption is unconfirmed. No major assistant has publicly committed to consuming it, and proof of impact is thin. It is low risk to add but do not expect guaranteed results.

Is llms.txt the same as robots.txt? No. robots.txt controls crawler access and is widely respected by compliant crawlers. llms.txt is a newer, curated content map with uncertain adoption.

Should I add an llms.txt file? If it costs about an hour, it is a reasonable low-risk bet, especially for docs-heavy sites. Prioritize extractable content, brand-fact consistency, and third-party mentions first.