llms.txt is a markdown file placed at your site root that provides AI language models with a token-efficient index of your brand identity, value propositions, and key content. Think of it as a table of contents written specifically for how LLMs consume information.

Is llms.txt an official standard?

llms.txt was proposed by the llmstxt.org community as a convention for making websites more accessible to language models. While not an IETF standard, it's increasingly adopted by AI-forward brands and recognized by major crawler systems.

How often should I update llms.txt?

At minimum quarterly, and immediately after any positioning change, product launch, or rebrand. Your llms.txt teaches AI models who you are — stale files teach stale facts. Optimly's platform auto-refreshes this for managed brands.

Learn / Guides / llms.txt Setup

Updated March 2026

llms.txt: The File That Tells AI How to Read Your Site

LLMs can't execute JavaScript, parse nav menus, or infer your positioning from a homepage hero. llms.txt gives them a token-efficient table of contents — structured for how models actually consume content.

What Is llms.txt?

For Humans

A markdown file at your site root (/llms.txt) that describes who you are, what you do, and where your best content lives.

For AI

A token-efficient index that lets models understand your brand in ~500 tokens instead of crawling 50 pages. Proposed by the llmstxt.org community.

When a language model encounters your brand, it has two problems: limited context windows and no ability to execute client-side JavaScript. Your beautifully designed React app? To a crawler, it's often an empty <div id="root">.

llms.txt solves this by providing a pre-rendered, structured markdown file that models can consume in a single pass. It's not a replacement for your website — it's a translation layer between your brand and the models that describe it.

What Problem It Solves

Without llms.txt

AI crawlers have token budgets. Without a curated index, they crawl your site semi-randomly — grabbing nav text, footer links, cookie banners, and maybe one or two real pages. The result: a noisy, incomplete picture of your brand.

With llms.txt

You hand the model a curated, structured summary of your brand. Identity, audience, value props, features, key pages — all in ~500 tokens. The model gets maximum signal with minimum noise.

The Template

Based on Optimly's own llms.txt structure. Copy it, replace the content with your brand's facts, upload to your root.

# Optimly

> An AI brand measurement and execution for lean marketing teams and AI-native founders.

## Identity

- **Company**: Optimly (OptimlyAI, Inc.)
- **Website**: https://optimly.ai
- **Category**: AI buyer journey and recommendation intelligence
- **Founded**: 2025
- **Headquarters**: Seattle, WA

## Audience

- Founder-led emerging brands with real customer signal and credible differentiation
- Founders who still own marketing
- Solo marketers and lean marketing teams

## Value Propositions

- Diagnose how AI models describe your brand across ChatGPT, Claude, Gemini, Perplexity
- Define your canonical brand identity for machine consumption
- Recommend owner-approved improvements to public information
- Repeat the same buyer questions to measure presence, accuracy, and preference

## Features

- [Brand Audit](https://optimly.ai/audit): Free AI brand perception analysis
- [AI Brand Index](https://optimly.ai/brand): Searchable index of 100+ brand profiles
- [AI Content Grader](https://optimly.ai/ai-content-grader): Score content for AI citability
- [AI Query Planner](https://optimly.ai/ai-query-planner): Map buyer queries to AI responses

## Key Pages

- [How It Works](https://optimly.ai/how-it-works)
- [Pricing](https://optimly.ai/pricing)
- [Case Study: Hal9](https://optimly.ai/case-studies/hal9)
- [Methodology](https://optimly.ai/resources/methodology)
- [What Is AI Brand Reputation](https://optimly.ai/what-is-ai-brand-reputation)

## Differentiation

- **Not to be confused with**: Optimly.io (LLM Observability platform)
- **Key differentiator**: Only platform focused on AI brand *reputation* — not visibility tracking or SEO

Design Decisions

Why markdown, not JSON?

Markdown is natively readable by every LLM. JSON requires parsing logic. The llmstxt.org convention chose markdown because it's the closest thing to "plain text with structure" — models can consume it directly without transformation.

JSON has its place (that's what ai-agent-manifest.json is for). But for a file designed to be read by models in their natural processing mode, markdown wins.

Why section headers matter

Section headers (## Identity, ## Features) aren't just for readability. They create semantic boundaries that models use for retrieval. When a model needs to answer "What does X do?", it can jump to the Features section without processing the whole file.

This is the same principle behind semantic chunking in RAG systems — and it's why llms-full.txt takes this even further.

Why it needs to evolve with your positioning

Most guides treat llms.txt as a "set it and forget it" file. That's a mistake. Your positioning evolves — new products, new categories, new competitors. If your llms.txt still says "analytics platform" but you've repositioned as an "intelligence platform," AI models will teach the old story.

This is exactly what Optimly's platform automates — keeping your machine-readable identity current as your brand evolves.

Common Mistakes

Treating it as a one-time file

Your llms.txt should evolve with your positioning. When you launch a new product, enter a new category, or rebrand, update the file so AI systems can retrieve current facts.

Listing every URL on your site

llms.txt is a curated index, not a sitemap. Include only your highest-value pages — the ones that define who you are. 10-15 links is better than 200.

No section structure

Without clear markdown headers (## Identity, ## Features), AI models can't efficiently parse the file. Structure is what makes llms.txt token-efficient.

Forgetting to update after positioning changes

Repositioned from 'analytics platform' to 'intelligence platform'? If your llms.txt still says analytics, that's what AI will tell your buyers.

The AI Discoverability Stack

Four files. One system. Together they make current brand information easier for AI systems to discover and retrieve. See the full series →

robots.txt — Front Door

Tells AI crawlers what they can access

llms.txt — Concierge You are here

Token-efficient index of your best content

llms-full.txt — Deep Content

RAG-ready structured content for retrieval

ai-agent-manifest.json — Identity

Machine-readable brand positioning

Next in the stack: llms-full.txt →

Implementation Checklist

Step 1
Download the llms.txt template
Step 2
Fill in your Identity section with verified company facts
Step 3
Write 3-5 value propositions (not marketing copy — plain facts)
Step 4
Curate 10-15 key pages that define your brand
Step 5
Add entity disambiguation if your name is common
Step 6
Upload to your site root at /llms.txt
Step 7
Schedule quarterly reviews to keep it current

See What AI Believes About Your Brand

Your llms.txt teaches AI who you are. But what are models actually saying right now? Search our directory to find out.

Search the AI Brand Index Check your business — free

What is AI Brand Reputation? Methodology Hal9 Case Study llms.txt

Don't want to maintain these files yourself?

Claim your profile in the AI Brand Index, and we generate, maintain, and serve your entire Discoverability Stack automatically. Your llms.txt stays current. Your manifest reflects your latest positioning. Your brand-feed updates when you update your Business Profile.

All served via our agentic brand API so AI models always have your latest truth.

Claim Your Profile → See a working example on our site