Most llms.txt generators ask an AI to describe your pages. This one doesn’t: it reads your sitemap and uses each page’s own title and meta description, so every line is something you already published. Enter your website, get a spec-valid llms.txt in seconds, then validate it with the companion checker — free, no signup.
Free · No signup · Built from your sitemap and your pages' own titles and descriptions
The generator reads your homepage, your robots.txt (for declared sitemaps), and your sitemap.xml — the same discovery path a crawler takes.
Up to 20 pages are chosen from your sitemap, shallowest paths first — the pages a human curator would pick — with asset files and duplicates skipped.
Every selected page contributes its own title and meta description. Nothing is invented or rewritten — if a page has no description, its line simply carries the title.
The composed file follows the llmstxt.org spec: an H1 title, a blockquote summary from your homepage, and H2 sections of link lists. Copy or download it, publish it at /llms.txt, then run the companion checker to confirm it scores 100.
H1 title, blockquote summary, and H2 sections with markdown link lists — the exact shape the llmstxt.org proposal defines, ready to publish as-is or edit first.
Pages are grouped by their URL structure — blog posts under Blog, docs under Docs — with top-level pages first in an Overview section.
Every description is the page’s own meta description. What you approved for search snippets is what AI assistants get — no hallucinated claims to proofread.
One-click copy or an llms.txt download, plus a direct handoff to the free llms.txt checker so you can confirm the published file scores clean.
Most generators send your pages to an AI model and publish whatever it writes. This one is deterministic: it reads your sitemap and uses each page’s own title and meta description, so every line is content you already wrote and approved. That makes the output instant, free of hallucinated claims, and safe to publish after a quick read.
Up to 20 pages from your sitemap (or just the homepage when no sitemap exists), preferring shallow paths — your homepage, pricing, product, and top-level docs before deep archive pages — and skipping images, scripts, and other asset URLs. The file is meant to be a curated index, not a mirror of your whole site.
Yes. A page without a description contributes a plain “- [title](url)” line, which is valid per the spec. But thin descriptions are worth fixing anyway: the same meta descriptions feed search snippets and AI answers alike.
At the root of your domain: https://yoursite.com/llms.txt. On most platforms that means dropping it in your public or static folder. After publishing, run the free llms.txt checker to confirm the live file is being served correctly — single-page apps often return their HTML shell for unknown paths, which reads as a missing file.
Adoption by AI providers is still emerging — no major assistant has committed to reading llms.txt the way search engines committed to sitemaps. It’s a low-cost, forward-looking signal: one file that gives assistants a curated map of your site. What you can measure today is whether AI mentions your brand, which is what the free AI visibility check does.
llms-full.txt is the spec’s companion file for full page content rather than a curated index. This generator deliberately produces only llms.txt — the index is the high-leverage part, and inlining full content is better done selectively once the index exists.
This check is one signal. The free audit grades your whole site for AI search and shows the plan Magister would run to fix what it finds.