What is llms.txt, and does it help with Google?
llms.txt is a plain-text file, served at your site's root, that gives AI systems a curated, machine-friendly summary of what your site is and where its most important content lives. Think of it as a reading guide for language models rather than a crawl directive like robots.txt. On the record: it does not help with Google. Google stated officially in June 2026 that its crawlers do not read llms.txt, so it has no effect on Google rankings or AI Overviews, and anyone selling it as a Google tactic is overclaiming. It remains a cheap, sensible courtesy for non-Google AI crawlers such as Perplexity's and Anthropic's, which can use it to orient on your site. Ship one because it costs minutes and may help some engines parse you; skip any vendor pitch that prices it as a ranking lever. This site publishes its own at /llms.txt.
What does llms.txt actually do?
It gives language-model crawlers a curated map: who you are, what the site covers, and which pages matter most, in plain markdown they can parse without crawling everything. The proposal emerged in 2024 as a lightweight convention, and some AI crawlers read it opportunistically. It carries no directives; it cannot block or admit anyone. Access control stays robots.txt's job.
What did Google actually say?
In June 2026 Google stated officially that its crawlers do not read llms.txt: no effect on rankings, no effect on AI Overviews inclusion. That closed a year of speculation and a fair amount of vendor overclaiming. If someone prices llms.txt into a proposal as a Google visibility lever, that line item is decoration.
Should you ship one anyway?
Yes, for three honest reasons. It may help non-Google engines (Perplexity's and Anthropic's crawlers among them) orient on your site faster. It costs minutes to generate from your sitemap and page descriptions, so the downside is nearly zero. And regenerating it on every deploy keeps you honest about which pages you actually consider important. This site's own file is at /llms.txt, generated from the same content collections that build the pages, so it never drifts from what is published.
What matters more than llms.txt?
Everything in the actual causal path: crawler access in robots.txt, Bing indexation, answer-first structure, freshness, and third-party citations. If your GEO budget is finite, spend it there.
Updated · Dreamware Development
Common questions
Is llms.txt the same as robots.txt?
No. robots.txt controls crawler access and is honored across the industry; llms.txt is an informational summary with no enforcement role. You need robots.txt correct first, including explicit allows for the AI crawlers you want, and llms.txt is an optional courtesy on top.
What belongs in an llms.txt file?
A one-line identity statement, a short summary of what the site offers, and a curated list of your most important pages with one-line descriptions. Keep it current; a stale llms.txt tells engines your site maintenance is stale too.
Why publish one if Google ignores it?
Because Google is not the only reader. Perplexity, Anthropic, and other AI crawlers can use it to orient, it costs minutes to generate, and keeping it honest signals operational care. Just never budget for it as a ranking tactic.
Keep reading
How do I show up in ChatGPT recommendations?
Five moves get a brand into ChatGPT's recommendations, allowing AI crawlers, Bing indexation, answer-first pages, freshness, and third-party citations. Here is the honest playbook.
AnswerWhat is generative engine optimization (GEO)?
Generative engine optimization (GEO) makes a brand visible, accurately described, and cited inside AI answers from ChatGPT, Perplexity, Gemini, and Google AI Overviews. Here is how it works.
Find out what AI says when your buyers ask
The $999 AI Visibility Audit measures your Share of Answer across six engines, shows where AI sends your buyers instead, and hands you a 90-day roadmap. Credited to month one if we work together.