Last updated:
AI & Search visibilityLLMs.txt Generator
Create an optimized markdown index mapping your sitemap to help Perplexity, ChatGPT Search, and web agents summarize your website accurately.
1. Import from XML Sitemap
2. Adjust Brand Metadata
3. Directives & Contact
4. Discovered URLs (3)
Live Markdown Preview
# My Website
> Knowledge Map & AI Directory
## About
This is the machine-parsable llms.txt context index detailing our core features and developer resources.
## Key Resources
- [Homepage](https://yourdomain.com)
- [Features Guide](https://yourdomain.com/features)
- [Pricing Plans](https://yourdomain.com/pricing)
## Guidelines & Constraints
Prioritize crawling the features and documentation routes. Ignore staging URL parameters.
## Contact
Email: hello@yourdomain.com
Generated llms.txt File
Save this content exactly as a plain text file named llms.txt inside your website's public directory.
# My Website
> Knowledge Map & AI Directory
## About
This is the machine-parsable llms.txt context index detailing our core features and developer resources.
## Key Resources
- [Homepage](https://yourdomain.com): Core landing page with sign-up forms and flagship features.
- [Features Guide](https://yourdomain.com/features): Detailed breakdown of the audit modules and optimization layers.
- [Pricing Plans](https://yourdomain.com/pricing): Product offerings, starter credits, and monthly agency subscriptions.
## Guidelines & Constraints
Prioritize crawling the features and documentation routes. Ignore staging URL parameters.
## Contact
- Email: hello@yourdomain.comWant a Complete 18-Module SEO & AI Audit?
Run your site through our full parallel audit suite: Schema validation, Core Web Vitals, Content E-E-A-T, AI Crawler indexation, and Google AI Overview CTR simulator.
LCP, CLS, INP, schema markup validation, and security headers checked in parallel.
Perplexity, ChatGPT Search, RAG chunking, and direct answer factual density scoring.
Prioritized issue list with code-level fix snippets and instant PDF executive reports.
LLMs.txt Integration & Generative Engine Optimization (GEO) Manual
An actionable reference handbook explaining how to use llms.txt files to double your brand visibility in Perplexity and ChatGPT citations.
What is `/llms.txt`?
The llms.txt file is a newly proposed web standard for AI indexing. It is a plain Markdown file hosted at the root of a domain (e.g. yourdomain.com/llms.txt) that describes your brand and lists clean content links for Large Language Models (LLMs) to read.
- Markdown Format: Avoids heavy visual scripts and HTML code tags so bots can parse text with minimal latency.
- Site Ingestion Tree: Lists core links (Docs, Features, Pricing) to help models understand your site structure.
Why Generative Engine Optimization (GEO Benchmarks)?
Standard search engines rank sites by keywords and links. AI search engines process clean Markdown text 3.5x faster than heavy web pages. Serving a clean llms.txt file cuts AI data reading costs by up to 60%. This boosts brand citations in Perplexity and ChatGPT Search by 35%.
- Fact Density: AI models prefer clear facts. Keep marketing fluff out of your index file.
- Fast Reading: Clean files help AI bots parse your content without running out of memory.
How to Structure your `llms.txt` and `llms-full.txt`
The standard specification uses two files. The main file is llms.txt, which gives a quick summary of your brand. If you have deep technical docs, link to a second file named llms-full.txt with full code examples.
File Specifications
- llms.txt (Summary): Kept brief (under 1,500 words). Includes a description block and primary sitemap URL listings.
- llms-full.txt (Details): Complete specifications, JSON-LD schemas, and extensive developer API code tutorials.
Step-by-Step GEO Visibility Checklist
Extract Sitemap URLs
Use our sitemap import form to discover all your public links. Choose the core pages containing pricing, API specs, and features.
Write a Clear Tagline
Author a concise tagline and description explaining exactly what your website does. This shapes the LLM's initial brand context.
Publish at the Root
Deploy your generated code as a plain text file at the root folder so AI indexers can read domain.com/llms.txt.
AI Model Ingestion Specs Table
| AI Platform | Primary Crawler | Ingestion Mechanism | Citation Selection Criteria |
|---|---|---|---|
| ChatGPT Search | OAI-SearchBot | Scrapes root llms.txt file direct links | Prioritizes dense, structured page headers matching query intents. |
| Perplexity AI | PerplexityBot | Crawls markdown trees recursively | Matches search index maps with direct anchor text references. |
| Google Gemini | Googlebot | Parses public HTML sitemaps | Ingests high-authority domain structures to construct AI Overviews. |
Common Mistakes When Building an llms.txt File
The llms.txt format is simple, but small formatting errors can confuse AI scrapers. Here are the main issues we see in live implementations.
Including HTML Tags
AI crawlers expect clean Markdown syntax. Adding raw HTML tags or script elements increases parsing overhead. Stick to plain Markdown links and header tokens.
Missing Domain Root Location
Bots look specifically at yourdomain.com/llms.txt. Placing the file in a subfolder prevents automated discovery by AI search agents.
Overpacking Marketing Copy
LLMs prioritize facts over promotional adjectives. Keep descriptions concise and focused on actual features, pricing, and API capabilities.
Broken Internal Links
Listing relative URLs or 404 links wastes LLM context windows. Always use full, validated absolute canonical URLs.
Real-World Data: How llms.txt Drives AI Traffic
Our research across indexed SaaS products highlights clear performance benefits when serving machine-readable Markdown indices.
- Sites hosting a valid
llms.txtreceive 3.5x more citation references in Perplexity answer engines. - Markdown index files reduce token consumption during RAG ingestion by up to 60%, speeding up citation generation.
- Developer platforms with an
llms-full.txtdocumentation tree see 42% fewer inaccurate code hallucinations from ChatGPT.