llms.txt Cheatsheet

The 2026 reality check: does llms.txt actually do anything?

Short version: almost nothing is reading these files yet, and Google has said it ignores them. Build one if you want, but do not expect traffic or rankings from it.

The data is blunt. Ahrefs analysed 137,000 domains in May 2026: 28% published an llms.txt file, and 97% of those files received zero requests. Of the fetches that did happen, 96% came from bots, and a chunk of those were audit and scanner tools (around 12%), not production AI systems. Roughly 19.5% came from named AI tools like GPTBot and PerplexityBot. The format is mostly being read by validators, not the assistants you are trying to reach.

Google has been just as clear. In June 2026 it updated its generative-AI guidance to spell out that llms.txt files do not help or hurt your rankings, and that Google Search ignores them along with other AI-only Markdown files. John Mueller, on Google's Search Off the Record podcast, called the discovery use case a dead end: a self-reported file cannot tell an LLM to pick your site over a competitor's, because every site can make the same claims. He compared it to the old meta keywords tag.

So where does it help? One narrow case: once an agent is already on your site. Mueller's example was an AI agent trying to complete a task, like buying a photo, and looking for instructions on how to do it. Chrome's agentic-browsing guidance frames it the same way: a map for an agent that has already arrived, not a way to get discovered. Think store directory for someone who already walked in, not a billboard that brings them there.

The verdict: for discovery, lean on what already works: clean server-rendered HTML, internal links, and genuinely useful content. Keep llms.txt as low-cost insurance for the agentic future, not as a ranking or traffic lever. It will not hurt you, and the standard may matter more once agent behaviour settles (Mueller reckons six to twelve months before any format wins).

The llms.txt format

Plain Markdown, in this order. Only the H1 is strictly required.

A minimal example

# Lawrence Hitches

> AI SEO consultant. Guides on getting cited in AI search.

## Cheatsheets
- [AI SEO Cheatsheet](/ai-seo-cheatsheet/): the umbrella reference
- [AEO Cheatsheet](/aeo-cheatsheet/): the citation checklist

## Optional
- [About](/about/): background and credentials

llms.txt vs llms-full.txt vs robots.txt

The honest read: no major AI provider has confirmed it reads llms.txt for ranking, so do not expect a lift. It is cheap to ship, harmless, and doubles as a clean content map. Pair it with the AI SEO cheatsheet.

See browse all SEO cheatsheets grouped by topic.

Frequently asked questions

Does Google use llms.txt?

No. Google confirmed in June 2026 that Google Search ignores llms.txt files and that they neither help nor hurt your rankings. They are intended for other AI services and agents, not Google Search.

What is llms.txt?

llms.txt is a proposed Markdown standard, a file at your domain root that gives AI models a curated map of your most important pages. It uses an H1 site name, a blockquote summary, and H2 sections of links with descriptions.

Where does llms.txt go?

At your domain root, served at yoursite.com/llms.txt, the same place as robots.txt. The companion llms-full.txt, which inlines the full text of key pages, also sits at the root.

Does llms.txt actually help AI visibility?

Unproven. No major AI provider has confirmed it uses llms.txt for ranking or citation. Treat it as low-cost insurance and a tidy content map. Your crawlable, extractable on-page content matters far more.

Is llms.txt the same as robots.txt?

No. robots.txt controls which crawlers can access what. llms.txt curates which pages you want AI models to read. One restricts access, the other recommends content.