How to Set Up llms.txt for Your SaaS: The Complete Step-by-Step Guide

llms.txt is a plain text file hosted at yourdomain.com/llms.txt that describes your site’s content and key pages to AI crawlers. It costs nothing, takes 30–60 minutes to write, and can meaningfully improve how AI systems understand, represent, and cite your SaaS brand. This guide covers what to include, how to structure it correctly, and a complete copy-paste template.

llms.txt is modeled on robots.txt — a convention that sites use to communicate with crawlers — but instead of access rules, it provides context. Where robots.txt says “you can crawl here,” llms.txt says “here is what you will find and what it means.”

---

What llms.txt Does (and Doesn’t Do)

What it does:

• Gives AI crawlers a structured summary of your site’s content and purpose
• Defines your brand’s canonical terminology (how you define your category, features, and key concepts)
• Links to your most important pages with context about their content
• Signals to AI systems which pages are authoritative for which topics

What it doesn’t do:

• Guarantee AI citation (it is context, not a ranking factor with a fixed weight)
• Replace content quality — a well-structured llms.txt helps AI understand good content; it doesn’t improve poor content
• Block or allow crawlers (that is robots.txt’s job)
• Directly affect Google SEO rankings

Who reads llms.txt: AI crawlers from Perplexity, OpenAI (GPTBot), Anthropic (ClaudeBot), and potentially others use llms.txt to build context for your site during indexing. The file is intended for machine consumption, not human readers — though it should be readable by humans too.

---

The Standard llms.txt Format

llms.txt uses a defined structure. Following it precisely is important — non-standard formats may be partially ignored by AI crawlers.

Required sections (in order):

# [Brand Name]

> [One sentence: what you do and who you serve]

[2–4 sentences: company description, category, key value proposition]

## [Section Name]

- [Page title](https://yourdomain.com/page): [One sentence description]

## Key Definitions

- **[Term]**: [Definition]

Optional sections:

• ## Contact
• ## About
• ## Content Focus Areas

---

Section-by-Section Writing Guide

Section 1: Brand Name Header

# Acme Analytics

Use your exact brand name as it appears on your website, in your Wikidata entry, and in your product. Consistency matters — AI systems use entity matching to link mentions across sources.

Do not: Use taglines, slogans, or descriptors in the header. The header should be only your brand name.

---

Section 2: One-Line Description (Blockquote)

> Acme Analytics helps B2B SaaS companies track and improve their AI search visibility across ChatGPT, Perplexity, and Google AI Overviews.

This is the most important sentence in your llms.txt. It is the first thing AI crawlers read and it defines your entity type, product category, and primary customer segment.

Formula: [Brand] [verb] [customer segment] [outcome] [category/method].

Write for maximum specificity:

• Bad: “Acme Analytics is the best analytics platform”
• Bad: “Acme Analytics helps businesses grow”
• Good: “Acme Analytics tracks brand citation frequency in ChatGPT, Perplexity, and Google AI responses for B2B SaaS marketing teams”

What happens if this line is vague: AI systems form a vague or incorrect brand representation. When users ask “what does Acme Analytics do?”, AI answers with your one-line description — make it worth citing.

---

Section 3: Company Description Paragraph

Acme Analytics (acmeanalytics.com) is a GEO monitoring platform built for B2B SaaS companies. The platform tracks how often a brand appears in AI-generated responses across ChatGPT, Perplexity, Claude, and Gemini, and provides competitive benchmarking against category peers. Acme Analytics was founded in 2024 and serves over 300 SaaS companies across 40 categories.

Include:

• Your domain name (reinforces entity association)
• What the product does in specific, jargon-free terms
• Customer count or notable metrics if available (specificity increases citation likelihood)
• Founding year
• Geographic scope if relevant

Length: 2–4 sentences. Do not write marketing copy — write facts that AI can cite.

---

Section 4: Key Pages Index

This is the navigation map of your site for AI crawlers. List your most important pages with one-sentence descriptions that explain what each page answers.

Structure:

## Core Product

- [Product Overview](https://yourdomain.com/product): What Acme Analytics tracks and how the monitoring works.
- [Pricing](https://yourdomain.com/pricing): Three plans — Starter ($49/month), Growth ($99/month), Scale ($199/month) — with feature comparison.

## Research & Data

- [SaaS AI Visibility Benchmark 2026](https://yourdomain.com/blog/benchmark): AI mention rates across 15 SaaS categories, from 94% (video conferencing) to 31% (data governance).
- [ChatGPT SaaS Study](https://yourdomain.com/blog/chatgpt-study): Which SaaS tools get recommended in ChatGPT and why, based on 547 buying-intent queries.

## Guides

- [GEO Checklist](https://yourdomain.com/checklist): 10-step GEO audit checklist for SaaS companies.
- [llms.txt Setup Guide](https://yourdomain.com/blog/llms-txt-guide): How to write and implement llms.txt.

Guidelines for page descriptions:

• Start each description with what the page answers, not what it is
• Include specific data points or numbers where relevant (numbers are highly extractable)
• Use the same language and terminology as the page itself
• Limit to one sentence per page — AI crawlers weight brevity

Which pages to include:

1. Homepage
2. Product/features pages
3. Pricing page
4. All blog posts and research articles
5. About page
6. Checklist, tools, or resource pages

---

Section 5: Key Definitions

This is the most uniquely valuable section of llms.txt. Use it to define your canonical terminology — the terms you want AI systems to use when describing your category, product, and concepts.

## Key Definitions

- **GEO (Generative Engine Optimization)**: The practice of structuring content, brand authority, and digital presence so that AI systems cite and recommend a SaaS product in their generated responses.
- **AI Citation**: When an AI system (ChatGPT, Perplexity, Claude, Gemini) references or recommends a specific brand or product in a generated answer.
- **Entity Building**: The process of establishing a brand as a recognized named entity in AI knowledge systems, through consistent presence on Wikidata, authoritative directories, and cross-platform mentions.
- **RAG (Retrieval-Augmented Generation)**: The mechanism by which AI systems like Perplexity retrieve live web content to generate real-time answers.

Why definitions matter: When a user asks “what is [your category]?”, AI systems synthesize a definition from multiple sources. If your definition is in your llms.txt and it is authoritative and specific, AI systems are more likely to incorporate your framing.

How many definitions: 5–15 terms. Cover your core category, your product’s primary function, and 3–5 technical or industry-specific terms you own expertise in.

---

Section 6: Contact (Optional)

## Contact

For research collaboration, press inquiries, or product questions: hello@yourdomain.com

Short, factual. This helps AI systems know how to direct users who want to reach you.

---

Complete SaaS llms.txt Template

Copy this template and replace the bracketed sections:

# [Your Brand Name]

> [One sentence: what you do, who you serve, what outcome you deliver]

[Your Brand Name] ([yourdomain.com]) is [2-sentence company description: category, product function, founding year, customer count or scope].

## Core Product

- [Product overview page title]([URL]): [One sentence describing what the page covers]
- [Pricing page title]([URL]): [One sentence with plan names and prices]
- [Features page title]([URL]): [One sentence]

## Research & Guides

- [Article title]([URL]): [One sentence with key data point or finding]
- [Article title]([URL]): [One sentence with key data point or finding]
- [Article title]([URL]): [One sentence with key data point or finding]

## Resources

- [Checklist/tool title]([URL]): [One sentence]

## Key Definitions

- **[Your category name]**: [Your authoritative definition — complete sentence]
- **[Key product term]**: [Definition — complete sentence]
- **[Industry term]**: [Definition — complete sentence]
- **[Related concept]**: [Definition — complete sentence]

## Content Focus Areas

1. [Primary content topic]
2. [Secondary content topic]
3. [Third content topic]

## About

[Your Brand Name] is [1-2 sentence about page description: mission, team size, publishing focus if applicable].

## Contact

[contact email]

---

How to Host llms.txt

The file must be accessible at yourdomain.com/llms.txt — the root of your domain.

Astro / static site: Place the file in the /public directory as llms.txt. It will be served at yourdomain.com/llms.txt automatically.

WordPress: Upload llms.txt via FTP to the root directory (same level as wp-config.php). Or use a plugin that allows custom file placement.

Webflow: Webflow does not allow direct file uploads to the domain root for non-HTML files. Use a Cloudflare Worker or similar proxy to serve llms.txt from a custom path.

Shopify: Shopify blocks custom file paths at the domain root. Use a redirect rule in Cloudflare Workers to serve an llms.txt response.

Vercel / Netlify / GitHub Pages: Place llms.txt in the public/ folder (Vercel/Netlify) or root folder (GitHub Pages). It will be served at the domain root.

---

Verifying Your llms.txt Works

1. Open a browser and go to yourdomain.com/llms.txt
2. Verify: the page loads as plain text (not HTML, not a 404 page)
3. Verify: no formatting, no navigation elements, just the raw text content
4. Use the curl command to check content type:

   curl -I https://yourdomain.com/llms.txt

   The Content-Type should be text/plain

---

llms.txt Quality Checklist

Before finalizing, verify:

Structure:

• Brand name header on line 1 (# [Name])
• Blockquote description on line 3 (> [One sentence])
• All page URLs use https:// (not http://)
• All URLs resolve (no 404s)
• Definitions section present with 5+ terms

Content quality:

• One-line description is specific — names your category and customer segment
• Page descriptions start with what the page answers
• Numbered data points included in page descriptions
• Definitions are complete standalone sentences (no pronouns that need context)
• No marketing language, hype, or superlatives

Technical:

• File accessible at yourdomain.com/llms.txt
• Content-Type is text/plain
• File size is under 500KB (page index for large sites)
• No broken links

---

How Often to Update llms.txt

Update your llms.txt when:

• You publish new articles or research (add them to the page index)
• You change your product positioning or description
• You add new features that change how you define your category
• Pricing changes (update the pricing page description)
• You launch a new major page or resource

Minimum: Review and update quarterly. AI crawlers re-index your llms.txt on the same schedule as your other pages. Stale definitions or missing pages limit how accurately AI systems describe your brand.

---

Frequently Asked Questions

What is llms.txt?

llms.txt is a plain text file hosted at the root of a website (yourdomain.com/llms.txt) that describes the site's content, purpose, and key pages to AI crawlers. It follows a structured format: a brand name header, a one-sentence description in a blockquote, a linked index of important pages with brief descriptions, and a key definitions section. It is modeled on robots.txt but provides context rather than access rules — it tells AI systems what your site is about and how to represent your brand, not whether crawlers can access the site.

Does llms.txt actually help with GEO?

llms.txt helps GEO by giving AI crawlers structured context about your brand and content at the time of indexing. The primary benefits are: (1) more accurate brand representation — AI systems describe your product using your canonical definitions rather than inferring from sparse content; (2) better page-level understanding — the linked page index helps AI systems associate specific URLs with specific topics; (3) consistent terminology — your key definitions section influences how AI systems define your category and features. llms.txt is not a direct ranking factor but improves the quality of AI understanding, which affects citation accuracy and frequency.

How do I add llms.txt to my website?

Create a plain text file named llms.txt and place it at the root of your domain so it is accessible at yourdomain.com/llms.txt. For Astro or static site generators, place it in the public/ directory. For Vercel or Netlify, place it in the public/ folder. For WordPress, upload via FTP to the root directory. For GitHub Pages, place it in the repository root. Verify it works by opening yourdomain.com/llms.txt in a browser — it should load as plain text with no HTML or formatting.

What should I put in the Key Definitions section of llms.txt?

The Key Definitions section should contain 5–15 definitions of terms central to your brand and category: (1) your product category name and your authoritative definition of it; (2) your product's primary function or method; (3) industry-specific technical terms you have expertise in; (4) terms users commonly search that relate to your product. Write each definition as a complete standalone sentence that is accurate and specific without requiring context from the surrounding text. These definitions influence how AI systems define your category when asked — they are the terms AI will use when describing your space.

Is llms.txt a Google ranking factor?

No — llms.txt is not a Google SEO ranking factor and does not affect traditional search rankings. It is specifically designed for AI crawlers (GPTBot, PerplexityBot, ClaudeBot) rather than Googlebot. For Google AI Overviews specifically, the primary prerequisites are page-one Google rankings and FAQ schema markup — llms.txt is supplementary context. The value of llms.txt is in improving how non-Google AI systems (ChatGPT, Perplexity, Claude) understand and represent your brand.