Describe your site for the models.
Build a valid llms.txt from a few fields, or paste your sitemap.xml and turn every URL into a link. Preview, copy and download the file — nothing leaves your browser.
Project
Becomes the H1, the first line a model reads.
Emitted as a blockquote right under the H1. Keep it to one sentence.
Plain paragraph placed between the summary and the first section.
Sections
Each section becomes an H2. Links use the form - [Title](url): optional description.
The convention marks secondary links as Optional so a model short on context can stop before them.
Paste a sitemap.xml
The XML is parsed locally and every <loc> becomes a link in a new “Pages” section. No network request is made.
llms.txt output
Checks
What llms.txt is — and what it is not
llms.txt is a proposed convention, published at llmstxt.org in September 2024, for placing a short, machine-readable Markdown file at https://example.com/llms.txt. The idea is simple: a language model that wants to understand a site should not have to parse navigation menus, cookie banners and JavaScript. It should be able to read one small file that says what the project is and where the useful pages live.
Be clear about its status. llms.txt is not an official standard, it is not a W3C or IETF specification, and it is not a de facto one either. No major AI vendor has committed to reading it, most crawlers never request it today, and there is no ranking, citation or training guarantee attached to it. It is a convention with growing tool support and a plausible future, adopted because it is cheap and harmless — not because anyone is obliged to honour it.
The format, line by line
The file is plain Markdown. An H1 names the project. A blockquote immediately below gives a one-line summary. Then one H2 per section, each followed by a list of links in the form - [Title](https://example.com/page): optional description. An H2 named Optional may be placed last to hold secondary links that a model with a limited context window can safely skip. That is the whole specification: there is nothing to compile and no schema to validate against.
Frequently Asked Questions
Is llms.txt an official standard?
No. It is a proposed convention from llmstxt.org (September 2024), not a specification from a standards body, and no major AI vendor has committed to reading it. Publish one because it is cheap and useful as documentation, not because anything requires it.
Will publishing llms.txt get my site into AI answers?
There is no guarantee. Some developer tools and retrieval systems fetch it when they find it; most crawlers never request it. Whether you are crawled is decided by robots.txt, and whether you are indexed is decided by search engines.
Do I still need robots.txt and sitemap.xml?
Yes — they answer different questions. robots.txt states permission, sitemap.xml handles discovery, and llms.txt offers context. None of the three replaces another, and a site can publish any subset of them.
What belongs in the Optional section?
Secondary links a model can safely skip: archives, changelogs, legal pages, older articles, press kits. The convention puts this section last and labels it Optional so a model short on context can simply stop reading before it.
Should I write llms.txt by hand or generate it?
Either works, because the format is only Markdown. Generating is safer for large sites: this tool checks absolute URLs, duplicate links and empty sections, and it can turn an existing sitemap.xml into a first draft in one click. Review the result before publishing — a generator cannot know which page matters most.
Related tools
robots.txt generator for AI crawlers · All MousyTools
Hecho por MousyDev
Creo herramientas web que respetan tu privacidad: tus datos se procesan en tu propio navegador y nunca pasan por mis servidores.