llms.txt
llms.txt is a plain-text, Markdown-formatted file placed at a website's root (yoursite.com/llms.txt) that gives large language models a curated map of your most important pages, in clean prose stripped of HTML, navigation, and ads. It was proposed in 2024 by Jeremy Howard of Answer.AI as an open convention to help AI systems find and understand your key content.
What llms.txt actually is
Think of llms.txt as a table of contents written for machines. Instead of forcing an AI crawler to parse your full HTML site -- menus, scripts, cookie banners, and all -- you hand it a short Markdown file that says what your site is about and links to the pages that matter, often with clean Markdown versions of those pages. It is conceptually the AI-era cousin of robots.txt, but where robots.txt tells crawlers where they may *not* go, this GEO convention tells LLMs where the *good stuff* lives.
- Lives at the root path: yoursite.com/llms.txt
- Plain Markdown -- an H1 with your site name, a short summary, then linked sections of key URLs
- An optional llms-full.txt can inline the full content of those pages
- Curated by you, not auto-generated noise -- the point is signal over coverage
For the full implementation walkthrough -- file structure, examples, and how to generate one -- see the in-depth guide to the llms.txt file.
Does it actually help? Be honest
This is where founders need a straight answer. As of 2026, no major LLM provider -- OpenAI, Anthropic, Google, Perplexity, Meta -- has publicly confirmed that its production answer engine reads llms.txt as a ranking or retrieval signal. Audits repeatedly find that GPTBot, ClaudeBot, and PerplexityBot mostly skip the file and crawl your HTML directly. Adoption is growing on the publisher side, but provider support is unproven.
So the honest take: llms.txt is cheap, low-risk, and forward-looking, but it is not a magic GEO lever today. It complements -- never replaces -- the fundamentals that actually drive AI citations: clean, factual, well-structured content, schema markup, and topical authority. Ship llms.txt as insurance, not as your strategy.
How a managed growth team treats it
On a one-person team, llms.txt is exactly the kind of small, defensible task that quietly never gets done. Ceres is a managed AI growth team where an AI Growth Officer orchestrates 11 specialists, and the GEO Strategist handles emerging conventions like this one -- drafting your llms.txt, keeping it in sync as pages change, and weighing it against the citation work that moves the needle. You stay the boss: the specialist drafts and proposes, and any change that publishes to your site stays approval-gated until you say yes. Want a baseline first? The free GEO audit checks whether your site is even set up to be cited.
FAQ
- What is llms.txt?
- llms.txt is a Markdown file at your website's root (yoursite.com/llms.txt) that gives AI models a curated, clean-text map of your most important pages, so they can understand your site without wading through HTML, scripts, and ads. It was proposed in 2024 by Jeremy Howard of Answer.AI.
- Do ChatGPT and Perplexity actually use llms.txt?
- Not in any confirmed way as of 2026. No major provider -- OpenAI, Anthropic, Google, Perplexity -- has publicly committed to using llms.txt in its production answer engine, and crawler audits show their bots mostly fetch HTML directly. Treat it as low-cost insurance, not a guaranteed ranking signal.
- Is llms.txt the same as robots.txt or a sitemap?
- No. robots.txt sets crawl permissions and a sitemap lists every URL for traditional search indexing. llms.txt is a short, human-curated summary in Markdown that points LLMs to your best content and explains what your site is about -- it favors signal over completeness.
An AI growth team that runs this for you
Ceres is a managed AI marketing team — you approve what ships. 14-day free trial, from $19/month.