SEO & AI Search

What Is llms.txt? A Complete Guide to the AI Search File

Key takeaways

  • llms.txt is a plain Markdown file at your site root that tells AI models which pages matter most.
  • It complements robots.txt and sitemap.xml — it doesn't replace either.
  • The format is tiny: one # H1, an optional summary quote, and ## sections of links.
  • You can generate a valid file in under a minute with a free browser tool.

More and more people get their answers from ChatGPT, Claude, Perplexity and Google's AI Overviews instead of clicking through ten blue links. That shift created a new question for every site owner: how do I make sure AI systems find, understand and correctly cite my content? The simplest first step is a small text file called llms.txt.

What llms.txt actually is

llms.txt is a plain-text Markdown file placed at the root of a website — reachable at https://yoursite.com/llms.txt — that gives large language models (LLMs) a short, curated map of your most important content. Think of it as a reading guide written for AI: instead of forcing a model to crawl and guess, you hand it a tidy list of your key pages with one-line descriptions.

The idea was proposed in 2024 by Jeremy Howard of Answer.AI and has since been adopted by documentation sites, developer tools and content platforms. It sits alongside the older "robots" ecosystem rather than replacing it — which is the part most people get wrong, so let's clear that up first.

llms.txt vs robots.txt vs sitemap.xml

The three files answer three different questions:

  • robots.txt — "May you crawl this?" Access control for crawlers. It allows or blocks paths, and points to your sitemap. (You can validate yours with our free robots.txt analyzer.)
  • sitemap.xml — "Here is everything." An exhaustive machine-readable list of every URL you want indexed, used by classic search engines.
  • llms.txt — "Here is what matters, and why." A short, human-readable, curated selection of your best pages with context, written so an AI model can quickly understand your site.

A search engine wants completeness; a language model with a limited context window wants relevance. That's the gap llms.txt fills.

The llms.txt format (it's tiny)

The whole specification is intentionally minimal Markdown:

  • One # H1 line with your site or project name — the only required element.
  • An optional > blockquote with a one-line summary of the site.
  • One or more ## H2 sections — such as Docs, Tools or Optional — each containing a Markdown list of links: - [Title](URL): short description.

Here's a real, valid example:

# WorkIQ — Free Online Tools

> 100+ free, no-signup browser tools for developers, students and marketers.

## Popular Tools

- [JSON Formatter](https://workiq.xyz/tools/json-formatter.html): Format and validate JSON in your browser
- [QR Code Generator](https://workiq.xyz/tools/qr-generator.html): Create QR codes for any URL or text

## Optional

- [Blog](https://workiq.xyz/blog.html): Guides on finance, dev and productivity

The ## Optional section has a special meaning in the proposal: models may skip it when they're short on context space, so put your nice-to-have links there and your must-read pages in the sections above it.

Create your llms.txt in under a minute

Paste your key URLs and get a valid, spec-compliant llms.txt to copy or download — free, in your browser, nothing uploaded.

Open the llms.txt Generator →

How to create and install yours

  1. Pick 5–20 pages that best represent your site. Cornerstone guides, key product or docs pages, your about page. Curate — don't dump your whole sitemap in.
  2. Write a one-line description for each. Accurate descriptions are what help a model cite you correctly.
  3. Build the file. Write the Markdown by hand or use our free llms.txt generator, which formats sections, titles and descriptions for you.
  4. Upload it to your site root so it loads at /llms.txt, next to robots.txt and sitemap.xml.
  5. Keep it fresh. Update the file when you publish important new content — a stale guide is almost as bad as none.

Does it actually help?

Honest answer: it's early, and Google has not announced any ranking benefit. What llms.txt does today is practical rather than magical: AI crawlers and agents that support it get a clean, unambiguous summary of your site, which improves the odds that answer engines describe and cite your pages accurately. Combined with clean headings and structured data, it's part of what's now called GEO — Generative Engine Optimization: making your content easy for AI systems to consume, not just for classic crawlers to index.

The cost is fifteen minutes once. For a file this cheap to create, being early is a reasonable bet.

Frequently asked questions

Is llms.txt an official web standard?

No — it's a community proposal (2024) with growing adoption. The format is stable, simple, and safe to publish today.

Does it replace robots.txt or my sitemap?

No. robots.txt controls access, sitemap.xml lists everything for search engines, and llms.txt curates highlights for AI models. Keep all three.

Will it boost my Google rankings?

Not directly. Its value is accurate citations in AI assistants and answer engines — an increasingly real traffic source.

What is llms-full.txt?

An optional companion file containing expanded, full-text content for models that want more than the summary. Start with plain llms.txt; add the full version only if you need it.

Related tools

See our Disclaimer for how to use WorkIQ content and tools.