See why AI recommends
your competitors, not you.

FEATURES

Built for clarity. Loved by AI.

WHO IT'S FOR

Built for one site or fifty

Same scan, same fix list — priced for how many sites you're responsible for.

Questions

What is llms.txt?

llms.txt is a plain-text file placed at the root of a website that gives AI assistants a clean, curated map of its content. It follows the proposal at llmstxt.org. A companion file, llms-full.txt, inlines the full text of every page.

Why does my website need one?

Web pages are built for browsers, so navigation, cookie banners and scripts surround the real content. In a typical scan, 95-99% of a page's bytes are removed before any content remains. Publishing llms.txt is a low-cost bet on an open proposal; the technical hygiene our scan checks — schema, headings, broken links, thin content — matters for search and AI alike regardless of whether any given crawler reads the file.

Do AI companies actually read llms.txt?

Unclear, and we won't pretend otherwise. Google's Search Advocate has called it speculative rather than confirmed, and no major AI company documents it as a ranking or citation signal. It costs nothing to publish and follows an open proposal at llmstxt.org, but treat it as a bet, not a guarantee — which is why our report leans on the 26 measurable technical checks and the live recommendation test, not the file alone.

Where do I put the file?

At the root of your domain, so it loads at yoursite.com/llms.txt. On Next.js that means the public folder; on WordPress, public_html. Webflow and Shopify can redirect /llms.txt to a hosted copy instead.

Does an llms.txt file get me recommended by ChatGPT?

On its own, no. Assistants answer from what web search returns, so being recommended means appearing in the pages they read. llms.txt makes you readable once they arrive; the citation check shows which pages are standing in your way.

How reliable is the AI recommendation test?

We ask each question twice per assistant rather than once, because a single query is noisy — published research on repeated LLM queries found under 1% odds of an identical answer on a second try. We report the split (e.g. "1 of 2 runs") instead of a false-certain yes/no. The models we query are web-connected and built on the same technology as ChatGPT, Gemini and Perplexity, but they are not the literal consumer apps, which add memory and personalisation we can't reproduce — so treat results as a strong signal, not a guarantee of what any one person sees.

Do you respect robots.txt?

Yes. Every scan reads robots.txt first and honours the Disallow rules that apply to it. Pages you have disallowed are never fetched or included.

What does it cost?

Scanning is free and covers your 5 most important pages. A one-time $19 payment per site unlocks the full crawl, all four files, and five live citation checks.