Create an llms.txt file that tells AI assistants which pages of your site actually matter.
llms.txt is a markdown file published at the root of a site — /llms.txt — that hands large language models a curated map of your content instead of making them guess. The proposal, published at llmstxt.org by Jeremy Howard in September 2024, defines a deliberately rigid skeleton so it can be parsed without ambiguity: an H1 with the name of the site or project, an optional blockquote with a one-paragraph summary, then H2 sections whose bodies are bullet lists of markdown links, each one optionally followed by a colon and a short description. A section named "Optional" has a special meaning: those links can be skipped when the context window is tight.
This generator builds exactly that structure. You fill in the name, the summary and as many sections as you need, adding one row per link; the output is assembled as valid markdown with the blockquote applied line by line, square brackets in titles escaped so they cannot break the link syntax, whitespace collapsed to keep every entry on a single line, and bare domains promoted to https:// while relative paths starting with / are left untouched. Empty sections and links without a URL are skipped rather than emitted as broken markdown, and the finished file downloads as llms.txt ready to upload.
Is it worth publishing? The honest answer: llms.txt is a community proposal, not a standard any AI vendor has publicly committed to reading, and it will not by itself get you cited in an AI answer. What it does is cheap and low-risk — a single small file, already adopted by a growing list of documentation sites — and it forces a useful exercise: deciding which twenty pages actually explain your product. Treat it as a companion to robots.txt (which controls crawling) and sitemap.xml (which lists every URL for search engines), not as a replacement for either. The file is generated in your browser and never sent anywhere.
# Site name → > one-paragraph summary → ## Section → - [Link title](https://url): short description. Sections named "Optional" may be skipped by tools with a limited context window.
It gives AI assistants a short, curated list of the pages that best explain your site, in markdown they can read directly, instead of leaving them to crawl and guess from navigation-heavy HTML.
At the root of the domain, served as plain text at https://yoursite.com/llms.txt, the same place robots.txt lives.
No AI provider has publicly confirmed that it does. llms.txt is an open proposal with real adoption among documentation sites, and publishing it is cheap, but treat any claim of guaranteed AI traffic with suspicion.
sitemap.xml lists every URL for search engines, robots.txt says what may be crawled, and llms.txt curates the handful of pages worth reading first, with descriptions written for a model rather than a crawler.
It is a companion convention that inlines the full markdown content of the listed pages into one large file, instead of linking to them. This tool generates the index file, llms.txt.
No. The markdown is assembled on your device by JavaScript, and the download is produced locally.
Vai.la turns any URL into a short link with click statistics, QR Code and your own biolink.
Vai.la is not responsible for how the tools are used or for decisions made based on their results.