How to Add llms.txt to Your Website (Step-by-Step Guide)

Enter llms.txt— a plain-text file that lives at the root of your website, purpose-built for large language models. Think of it as robots.txt, but instead of telling bots where not to go, it tells AI agents exactly what your content is, where the good stuff lives, and how it should be understood.

This guide walks you through everything: what llms.txt is, why it matters in 2026, and the exact steps to add one to your site today.

What Is llms.txt?

The llms.txtstandard was proposed in late 2024 and has gained rapid adoption as AI-powered search and agents have gone mainstream. The idea is simple: as LLMs browse the web on behalf of users, they need a reliable, low-noise signal about your site's content and purpose.

A web page full of navigation menus, cookie banners, sidebar ads, and footer links is noisy for an AI to parse. llms.txtcuts through that — giving a clean, structured summary of what your site offers and linking to the most important content in markdown format.

It lives at https://yourdomain.com/llms.txt and follows a straightforward markdown-based format.

Why Does It Matter?

Here is what has changed: AI agents are becoming the first layer of discovery for millions of users. When someone asks ChatGPT “what's the best tool for X?” or uses Perplexity to research a purchase, those systems are crawling the web in real time. If your site does not have clear signals for AI agents, you risk being misunderstood, misrepresented, or skipped entirely.

Beyond search, think about:

• AI coding assistants that pull in documentation from your site
• Autonomous agents performing research tasks on behalf of users
• RAG pipelines that index your content to answer customer questions
• AI browser agents that need to understand your site's structure before interacting with it

Sites that speak the language of AI agents in 2026 will have a compounding advantage — showing up more accurately, more often, in more AI-powered contexts.

The llms.txt Format Explained

Before writing your file, let's understand the structure. An llms.txt file is written in Markdown and follows this anatomy:

# Your Site or Company Name

> A one or two sentence description of what your site/product does.
> This is the most important part — it's what the AI reads first.

Optional additional context paragraphs go here.

## Section Name

- [Page Title](https://yourdomain.com/page): Brief description.
- [Another Page](https://yourdomain.com/another): Brief description.

## Another Section

- [Docs Page](https://yourdomain.com/docs/getting-started): Getting started guide.
- [API Reference](https://yourdomain.com/docs/api): Full API reference.

## Optional

- [Terms of Service](https://yourdomain.com/terms): Optional.
- [Privacy Policy](https://yourdomain.com/privacy): Optional.

Key rules:

• Line 1: An H1 with your site/company name
• The blockquote (>): A concise description — this is the highest-signal section
• Sections (H2): Logical groupings of your content
• List items: Markdown links with a brief colon-separated description
• “Optional” section: Anything that is useful but lower priority for AI consumption

Step-by-Step: Adding llms.txt to Your Website

Step 1: Plan Your Content Map

Before writing a single line, open a blank document and answer these questions:

1. What does my site/product actually do in one sentence?
2. What are the 5–10 most important pages an AI should know about?
3. How should those pages be grouped?

For a developer tool, your sections might be: Docs, API Reference, Guides, Blog. For an e-commerce site: Products, Categories, About. For a SaaS: Features, Pricing, Documentation, Support.

Step 2: Write Your llms.txt File

Here is a real-world example for a developer documentation site:

# Acme Dev Tools

> Acme Dev Tools provides a REST API and SDKs for adding real-time
> collaboration features to web applications. Trusted by 12,000+ developers.

Our tools handle presence indicators, cursor sharing, live cursors,
and conflict-free document editing (CRDT-based). We have SDKs for
JavaScript, Python, Go, and Rust.

## Getting Started

- [Quick Start Guide](https://acme.dev/docs/quickstart): Install the SDK and build your first real-time feature in under 10 minutes.
- [Core Concepts](https://acme.dev/docs/concepts): Understand rooms, presence, and document sync.
- [Authentication](https://acme.dev/docs/auth): How to authenticate users and issue JWT tokens.

## API Reference

- [REST API](https://acme.dev/docs/api): Full REST API reference with request/response examples.
- [JavaScript SDK](https://acme.dev/docs/sdk/js): Complete JavaScript/TypeScript SDK reference.
- [Webhooks](https://acme.dev/docs/webhooks): Configure webhooks to react to room events.

## Guides

- [Building a Collaborative Editor](https://acme.dev/guides/collaborative-editor): Step-by-step guide.
- [Presence Indicators](https://acme.dev/guides/presence): Show who's online.
- [Conflict Resolution](https://acme.dev/guides/crdt): How our CRDT implementation handles concurrent edits.

## Pricing & Plans

- [Pricing](https://acme.dev/pricing): Free tier, Pro ($49/mo), and Enterprise plans.

## Optional

- [Status Page](https://status.acme.dev): Real-time API uptime and incident history.
- [Changelog](https://acme.dev/changelog): What's new in each release.
- [GitHub](https://github.com/acme/acme-js): Open source SDK repository.

Step 3: Create the File

Save your content as a plain text file named llms.txt. No HTML, no special encoding — just UTF-8 plain text written in Markdown format.

Step 4: Serve It at the Right URL

The file must be accessible at https://yourdomain.com/llms.txt. How you do this depends on your stack:

Static Sites (Netlify, Vercel, GitHub Pages)

Simply place llms.txt in your public/ or root directory. It will be served automatically.

Next.js

Place it in your /publicdirectory — Next.js serves everything in /public as static files.

Express.js

const express = require('express');
const path = require('path');
const app = express();

// Serve llms.txt as static
app.use(express.static(path.join(__dirname, 'public')));

// Or serve it explicitly
app.get('/llms.txt', (req, res) => {
  res.setHeader('Content-Type', 'text/plain');
  res.sendFile(path.join(__dirname, 'llms.txt'));
});

WordPress

Upload llms.txt via FTP or your hosting file manager to the root of your WordPress installation (same directory as wp-config.php).

nginx

server {
    # ... your existing config ...

    location = /llms.txt {
        root /var/www/html;
        default_type text/plain;
    }
}

Step 5: Verify It Is Live

Open your browser and navigate to https://yourdomain.com/llms.txt. You should see your plain text file. If you see a 404, double-check your file placement and server config.

You can also run a quick curl check:

curl -I https://yourdomain.com/llms.txt
# Should return: HTTP/2 200
# Content-Type: text/plain

Step 6: Add the llms-full.txt Variant (Optional but Recommended)

The llms.txt standard also supports a companion file: llms-full.txt. While llms.txt is a directory/index (links to content), llms-full.txt contains the actual full content of your most important pages, pre-processed into clean markdown.

This is especially powerful for documentation sites, where you want AI agents to have the complete content without having to crawl each page individually.

Common Mistakes to Avoid

• Listing every single page.Keep it to your 10–20 most important pages. AI agents do not need your tag pages, author archives, or paginated lists.
• Vague descriptions.“About page” is useless. “Company background, team, and mission statement for Acme Dev Tools” is useful.
• Forgetting to update it. Set a reminder to review your llms.txt quarterly. When you add major features or new docs sections, update the file.
• Wrong content type. Serve it as text/plain, not text/html. Some AI crawlers are picky about this.
• Skipping the blockquote description. The > description is the highest-weight signal in the file. Do not leave it out or make it generic.

What Comes After llms.txt?

llms.txt is a great start, but it is one piece of a larger AI-readiness puzzle. To be fully optimized for the way AI agents interact with websites in 2026, you also need:

• Structured data / JSON-LD so AI agents understand your content type and entities
• Clean semantic HTML that parses well without a browser
• Descriptive alt text on images (AI vision agents read these)
• Machine-readable pricing and feature data
• A well-formed robots.txt that does not accidentally block AI crawlers