Generate a clean llms.txt for your site — and watch it get written.
Trakkr maps your site, reads the key pages, and writes a spec-compliant llms.txt you can copy or download. The file AI models read to understand you.
Free · No signup · About 15 seconds
Or try one of these
An llms.txt is a small Markdown file at your site's root — /llms.txt — that tells AI models like ChatGPT, Claude and Perplexity what your site is and where its important content lives. Think of it as a clean map written for machines, not a sitemap of every URL.
It's an open spec from llmstxt.org. In May 2026 Google added it to Chrome Lighthouse's Agentic Browsing audit — so it's quickly becoming a baseline signal of an AI-ready site.
# Your BrandAn H1 with the site name — the one strictly required line.> What the company is and does.A blockquote summary: plain facts, not a tagline.## ProductNamed sections that mirror your real site structure.- [Pricing](/pricing): plans and limits.Each link with a specific one-line description.## OptionalSecondary pages a model can skip under tight context.
- Discover
Map the site
Reads your homepage, sitemap, robots.txt and any existing llms.txt to find the pages that matter.
- Enrich
Read the key pages
Fetches the highest-signal pages for their real titles and descriptions — so link summaries are grounded, never invented.
- Compose
Write llms.txt
Composes a clean, spec-compliant file with grouped sections and concrete descriptions, using only URLs it actually found.
An llms.txt is a Markdown file served at your site's root (yourdomain.com/llms.txt) that gives AI models a structured map of what your site is and where its key content lives. It's an open standard from llmstxt.org, and in May 2026 Google added it to Chrome Lighthouse's Agentic Browsing audit.
Yes — completely free with no signup. Enter any domain and watch a spec-compliant llms.txt get written in about fifteen seconds, then copy or download it. There is a gentle limit of ten builds an hour per visitor to keep the tool fast for everyone.
Trakkr fetches your homepage, sitemap and robots.txt, reads your highest-signal pages for their real titles and descriptions, then composes the file. It only uses URLs it actually found on your site — it never invents or guesses links — so what you get is safe to publish as-is.
Upload llms.txt to your site root so it is served at https://yourdomain.com/llms.txt. On Vercel, drop it in /public; on Netlify, your publish directory; on WordPress, the site root via SFTP. The completed page has a short host-by-host guide.
It helps models understand your site cleanly, which is a useful baseline. But an llms.txt is a snapshot — it goes stale as your site changes, and publishing it does not tell you whether ChatGPT, Claude or Perplexity actually read or recommend you. Trakkr keeps the file current and tracks whether AI models act on it.
If a site blocks automated crawlers we build a starter file from whatever is public, and tell you clearly that discovery was limited. You can paste a sitemap URL or connect your site in Trakkr for the full map — what's produced is still a valid file you can publish.