How to Write an llms.txt File (Free Generator Included)

AI assistants like ChatGPT, Claude, and Perplexity don't crawl your site the way Google does. They need a fast, structured map of what your site contains and where the content that matters actually lives. That's what an llms.txt file is for — and if you don't have one yet, you're making AI crawlers work harder than they should to understand and cite your brand.

This guide walks through exactly how to write an llms.txt file by hand, then shows you the faster route: our free llms.txt generator, which crawls your sitemap and builds the file for you in seconds. For the fuller conceptual breakdown, see What Is llms.txt and Why Does It Matter?

The Standard llms.txt Format

The accepted structure follows a simple Markdown convention: an H1 with your site name, a blockquote summary, then H2 sections grouping links with short descriptions.

# Site Name

> One or two sentence summary of what the site/company does.

## Docs

- [Getting Started](https://example.com/docs/getting-started): Setup and installation guide

## Guides

- [Best Practices](https://example.com/guides/best-practices): Recommended configuration patterns

## Optional

- [Changelog](https://example.com/changelog): Release notes and version history

1. Start with an H1 containing your site or company name.
2. Add a blockquote summary immediately after — the single most important line for AI comprehension.
3. Group links under H2 sections that reflect your actual site structure.
4. Use a colon plus short description after each link — this is what the model uses to decide relevance.
5. Keep an "Optional" section for lower-priority pages the model can skip.

Step-by-Step: How to Write Your llms.txt File

1. Audit your most important pages. List the 15–40 URLs that best represent your product, services, and expertise. Skip login pages, legal boilerplate, and thin utility pages.

2. Group them logically. Mirror how a new visitor — or an AI model researching your category — would want the information organized: Docs, Guides, Services, Blog, Case Studies, Pricing.

3. Write one-line, specific descriptions. Avoid vague copy like "Learn more." Write the way you'd explain the page to a colleague in one breath.

4. Add a strong summary blockquote. This is the line most likely to be quoted or paraphrased by an AI assistant summarizing your brand.

5. Publish at the root. Save the file as llms.txt and serve it at https://yourdomain.com/llms.txt — not nested in a subfolder.

6. Keep it current. Treat llms.txt like a sitemap: revisit it whenever you ship major new content or restructure navigation.

Skip the Manual Work: Use the Free Generator

Writing this by hand for a site with hundreds of pages is tedious, and it's easy to miss sections or let the file go stale. Our free llms.txt generator does the heavy lifting: enter any domain, it crawls the sitemap automatically, groups content by section, and gives you a ready-to-use file to copy or download instantly. No signup required.

Common Mistakes to Avoid

• Dumping your entire sitemap in. llms.txt should be curated, not exhaustive — that's what sitemap.xml is for.
• Generic descriptions. "Home page" or "Learn more" tell an AI model nothing useful.
• Forgetting to update it. A stale llms.txt pointing to deprecated pages actively hurts AI comprehension.
• Treating it as a ranking hack. It improves comprehension and citation quality — it isn't a guaranteed ranking signal on its own.
• Skipping the summary blockquote. This single sentence often carries more weight than the entire link list.

How This Fits Into a Broader GEO Strategy

llms.txt is one piece of a much larger discipline: Generative Engine Optimization. Getting cited consistently by ChatGPT, Perplexity, and Google AI Overviews also depends on structured data, clear entity signals, authoritative content, and a site architecture that's easy for models to parse. Our GEO services team builds and maintains this infrastructure for growing brands.