Does Your Website Have An LLMs.txt?

LLMs.txt is a single Markdown file that lives at /llms.txt. Inside you’ll find a title, a short summary of your brand, and your most important pages sorted into sections, each with a quick description. Think of it as a reading list you hand to AI. Rather than crawling your whole site and guessing what matters, a model gets a clear starting point from you.

If you want agents and AI tools that visit your site to understand it quickly and accurately, it’s worth having. Left alone, these tools build their picture of you from whatever they happen to crawl, so your best pages often get skipped and old descriptions stick around. An LLMs.txt is the cheapest way to point them at the pages you’d actually choose.

Yes, and it’s growing fast. Adoption has roughly doubled or more year over year, and roughly one in ten sites now publish one. It’s not a ranking lever for Google Search or AI Overviews; Google’s said directly that it doesn’t affect those. But agent frameworks and coding tools (Claude Code, Cursor, OpenAI’s Agents SDK) do read it, and Chrome’s Lighthouse now audits sites for it as part of agentic-browsing readiness. No format is universal yet, but adding yours takes minutes and puts you ahead of the sites still sitting on the fence.

They solve different problems. robots.txt sets rules for what crawlers may touch. A sitemap lists every URL you have so search engines can index them. An LLMs.txt is the curated short version for language models: only the pages worth reading, each with a note on what it covers, so a model doesn’t end up describing your product from your careers page.

After downloading, add it to the root of your domain, so it loads at yoursite.com/llms.txt, right next to your robots.txt. It’s a plain static file, so there’s no build step and no code change. If you can upload a file to your site, you can publish this in a couple of minutes.