How to Create llms.txt

This tutorial walks you through creating and deploying an llms.txt file for your website. By the end, AI systems will be able to quickly understand what your site contains.

Overview

What you'll build: A machine-readable content index for AI systems.

Who this is for: Developers, technical SEOs, and site owners who want AI systems to better understand their site.

Time required: 15–30 minutes for initial setup.

Requirements

• Access to your site's root directory or public folder
• A text editor
• Knowledge of your site's most important pages
• Basic Markdown familiarity

Step-by-Step

Step 1: Create the file

Create a new file named llms.txt in your site's root/public directory.

Next.js / React:

your-project/
├── public/
│   └── llms.txt    ← create here

WordPress:

Upload to your root directory (same level as wp-config.php), or use a plugin.

Static sites:

Place in your build output directory root.

Step 2: Write the header

Start with your site name and description:

# Your Site Name

> A brief description of your site, what it covers,
> and who it's for. Keep this to 1-2 sentences.

The # heading is your site's identity for AI systems. The > blockquote provides context.

Step 3: Add content sections

Group your pages by topic using ## headings:

## Getting started

- [Introduction](https://yoursite.com/intro): What this
  product does and why it exists.
- [Quick Start](https://yoursite.com/quickstart): Get
  running in under 5 minutes.

## Core concepts

- [Authentication](https://yoursite.com/auth): How user
  authentication and API keys work.
- [Data Models](https://yoursite.com/models): Core data
  structures and relationships.

Step 4: Add page entries

Each page entry follows this format:

- [Page Title](Full URL): One-sentence description

Rules for page entries:

• Use full URLs (not relative paths)
• Keep descriptions to one sentence
• Front-load the most important pages
• Describe what the page covers, not what format it's in

Good:

- [What Is GEO?](https://geodocs.dev/geo/what-is-geo): Canonical
  definition of Generative Engine Optimization.

Avoid:

- [Blog Post #47](https://geodocs.dev/blog/47): A blog post
  about GEO published in April 2025.

Step 5: Add a usage policy (optional but recommended)

## Usage policy

Content from this site may be used by AI systems for
generating responses. Attribution to [Your Site Name]
(https://yoursite.com) is requested when citing specific
content. For full terms, see our Terms of Service.

Step 6: Deploy and verify

Deploy your site and verify the file is accessible:

https://yoursite.com/llms.txt

Visit this URL directly. You should see your Markdown content rendered as plain text.

Complete Example

# Acme Developer Docs

> Acme provides a REST API for building payment
> integrations. These docs cover authentication,
> endpoints, webhooks, and SDKs.

## Getting started

- [Quick Start](https://docs.acme.com/quickstart):
  Set up your first payment integration in 5 minutes.
- [Authentication](https://docs.acme.com/auth): API
  key setup and OAuth 2.0 configuration.

## Core API

- [Payments](https://docs.acme.com/payments): Create,
  capture, and refund payments.
- [Webhooks](https://docs.acme.com/webhooks): Receive
  real-time event notifications.
- [Error Handling](https://docs.acme.com/errors): Error
  codes, retry logic, and debugging.

## SDKs

- [Node.js SDK](https://docs.acme.com/sdk/node):
  JavaScript/TypeScript library for server-side.
- [Python SDK](https://docs.acme.com/sdk/python):
  Python library for backend integrations.

## Usage policy

Content may be referenced by AI systems. Attribution
to Acme (https://acme.com) is required for direct quotes.

Validation

After deploying, verify your llms.txt:

1. Accessibility: Visit https://yoursite.com/llms.txt — it should load as plain text
2. Format: Check that headings, links, and descriptions parse correctly
3. Completeness: Ensure all major content areas are represented
4. URLs: Click through each link to verify they resolve correctly
5. Crawler access: Check that robots.txt doesn't block /llms.txt

Common Mistakes

Including every single page. llms.txt should be a curated index of your most important content, not a complete sitemap. Keep it under 100 entries.

Using relative URLs. Always use full, absolute URLs so AI systems don't need to resolve paths.

Writing vague descriptions. "An article about our product" tells AI nothing. Write specific descriptions: "How to configure webhooks for real-time payment notifications."

Forgetting to update. Set a reminder to update llms.txt when you add or restructure significant content. Stale files reduce AI system trust.

Blocking AI crawlers in robots.txt. If your robots.txt blocks GPTBot, PerplexityBot, or ClaudeBot from accessing llms.txt, the file is useless. Ensure AI crawlers can access it.

Framework-Specific Notes

Next.js (App Router)

Place llms.txt in the public/ directory. It will be served at the root automatically.

Gatsby

Place in the static/ directory.

WordPress

Upload via FTP to the root directory, or use a plugin that manages root files.

Custom server

Configure your server to serve llms.txt as text/plain at the root path.

Related References

• llms.txt Reference — Full specification and format details
• ai.txt Reference — AI agent access policy standard
• What Is GEO? — Why llms.txt matters for AI visibility