Build a spec-compliant llms.txt file for your website in under a minute. Fill in your site details, add your key pages, then copy or download the file and drop it at your web root.
Follows the llmstxt.org proposal. Everything runs in your browser: nothing is uploaded or stored.
Group your most important URLs into sections. Use "Optional" for secondary pages LLMs can skip when context is tight.
# Acme Analytics > Acme Analytics is a product analytics platform that helps SaaS teams understand user behavior and reduce churn. Acme serves 2,000+ B2B customers. The docs below cover setup, APIs, and integrations. ## Docs - [Quickstart](https://acme.com/docs/quickstart): Install the tracker and send your first event in 5 minutes - [API Reference](https://acme.com/docs/api): REST endpoints, authentication, rate limits ## Blog - [Churn benchmarks 2026](https://acme.com/blog/churn-benchmarks): Original data from 1,400 SaaS companies ## Optional - [Pricing](https://acme.com/pricing) - [About](https://acme.com/about)
Upload this file to your web root so it resolves at yoursite.com/llms.txt. See ours live at geonimo.com/llms.txt.
llms.txt is a proposed web standard: a markdown file at the root of your site that gives large language models a concise, curated map of your most important content. Context windows are limited and full HTML pages are noisy, so llms.txt hands AI systems exactly what matters, in a format they parse natively.
The format is deliberately simple. An H1 with your site name, a blockquote with a one-line summary, optional context paragraphs, then H2 sections listing your key pages as markdown links with short descriptions. A special ## Optional section marks pages an LLM can skip when it needs to save tokens.
The spec is public and short: proposed by Jeremy Howard of Answer.AI on September 3, 2024, and maintained in the open at AnswerDotAI/llms-txt. Two details from the spec worth knowing before you write one: the H1 site name is "the only required section", and the authors position the file as an inference-time aid ("mainly useful for inference", in their words), not a training opt-out. llms.txt does not replace robots.txt and does not opt your content in or out of anything.
An honest note on adoption: publishing is widespread, consuming is not. The spec lists VitePress, Docusaurus and Drupal integrations, and Anthropic and Perplexity ship llms.txt files for their own docs, but no engine has publicly committed to reading the file, and Google's John Mueller compared it to the keywords meta tag in 2025. We still think it is worth five minutes: the cost is near zero, choosing your key pages is a useful exercise in itself, and the file is ready the day any engine starts consuming it.
Publishing an llms.txt pairs well with the other technical check: whether AI crawlers can access your site at all. Run our free AI crawlability checker to find out in seconds.
From generated file to live standard in a couple of minutes.
Place the file in your public or static folder so it deploys to yoursite.com/llms.txt. On Vercel, Netlify or any static host, that is a one-file commit.
Open yoursite.com/llms.txt in a browser. You should see raw markdown, served as plain text. If your router rewrites unknown paths, add an exception for the file.
Revisit it when you ship major pages. A stale llms.txt pointing at dead URLs is worse than none: it teaches models the wrong map of your site.
llms.txt is a plain-markdown file placed at the root of your website (yoursite.com/llms.txt) that gives large language models a curated, token-efficient map of your most important content. Proposed by Jeremy Howard of Answer.AI in September 2024, it uses a simple format: an H1 with your site name, a one-line summary quote, optional context, and H2 sections listing your key pages as markdown links.
Publisher-side adoption is real: the spec lists VitePress, Docusaurus and Drupal integrations, and Anthropic and Perplexity publish llms.txt files for their own docs. But no major AI engine has publicly committed to consuming it, and Google's John Mueller said in 2025 that no AI service has said they use it, comparing it to the keywords meta tag. Publish one because the cost is near zero and it doubles as a curated content map, not because it guarantees visibility.
robots.txt controls which crawlers may access which paths. sitemap.xml is an exhaustive machine-readable list of every URL for search indexing. llms.txt is different: it is a curated, human-written guide that tells an LLM which pages matter most and what your site is about, in markdown the model can ingest directly.
llms-full.txt is a companion convention: instead of linking to your key pages, it inlines their full content into one large markdown file so an LLM can read everything in a single request. It is most common for documentation sites. Start with llms.txt first; add llms-full.txt if your docs are the product.
Upload it to your web root so it resolves at https://yoursite.com/llms.txt, served as plain text (UTF-8). On most stacks that means dropping the file in your public or static folder and deploying. Then open the URL in a browser to verify it renders as raw markdown.
It is one signal among many, not a silver bullet. llms.txt makes your best content easier for AI systems to find and ingest, but citations in ChatGPT, Perplexity or Google AI are earned through citable content, third-party mentions, structured data and authority. Treat llms.txt as cheap, sensible hygiene inside a broader GEO strategy.
Now find out what ChatGPT, Perplexity, Claude and Google AI actually say about your brand, and what to fix to get cited. Free audit, first insights within 24 hours.