What Is LLMs.txt?
LLMs.txt is a Markdown file placed at your website's root (e.g., `llmrelevance.com/llms.txt`) that tells AI systems like ChatGPT, Claude, and Perplexity which pages matter most on your site.
Think of it as a curated table of contents for AI crawlers.
The Problem It Solves
Modern websites are hard for AI to read:
1. JavaScript-Heavy Sites
Most AI crawlers only read basic HTML. If your content loads via JavaScript, they might miss it entirely.
2. Too Much Noise
Navigation menus, ads, cookie banners, and sidebars create clutter. AI systems waste resources parsing irrelevant content.
3. No Clear Priorities
AI doesn't know which pages are your best resources versus outdated blog posts from 2018.
LLMs.txt solves this by providing a clean, structured guide to your most valuable content.
How It Works: The Technical Details
File Location
Place `llms.txt` at your website's root directory:
Format: Markdown
The file uses Markdown (the same format as GitHub README files), which AI systems can easily parse.
Basic Structure:
```markdown
Company Name
> One-sentence description of what you do
Products
- [Product 1](https://example.com/product-1): What it does and why it matters
- [Product 2](https://example.com/product-2): Key benefits in 10-15 words
Documentation
- [Getting Started](https://example.com/docs/start): How to begin using our platform
- [API Reference](https://example.com/api): Complete API documentation with examples
Company
- [About Us](https://example.com/about): Team, mission, and background
- [Contact](https://example.com/contact): How to reach our support team
```
Our Implementation
We created an llms.txt file for LLM Relevance Directory. View it here β
Our file includes:
- Core platform pages (Tools Directory, AI Search Playbook, Learn Hub)
- 20+ featured tools organized by category
- Educational content (comparison articles, strategy guides)
- Vendor resources (Submit Tool, Pricing, Partnerships)
Current Adoption: The Real Numbers
According to NerdyData research, as of July 2025, only 951 websites had implemented llms.txt.
That's a tiny fraction of the internet.
Who's Using It?
Developer-Focused Companies:
- Hugging Face - Comprehensive docs with code examples
- Vercel - Step-by-step AI SDK documentation
- Zapier - Simple list of API docs
- Cal.com - Minimalist link collection
- Anthropic - Claude's own documentation (interesting!)
Notable: Anthropic (makers of Claude) has implemented llms.txt on their own site. This doesn't mean Claude uses these files yet, but it suggests they're considering it.
Why Such Low Adoption?
No Official Support (Yet)
Google's John Mueller confirmed on Bluesky: *"No AI system currently uses llms.txt."* ([Source](https://www.semrush.com/blog/llms-txt/))
OpenAI, Google, and Anthropic haven't made public statements about reading these files when crawling websites.
Experimental Status
This is a *proposed* standard, not an adopted one. Early adopters are essentially betting on future adoption.
Should You Implement LLMs.txt? (Decision Framework)
β Consider It If You:
1. Already Optimize for AI Visibility
If you're tracking brand mentions in ChatGPT with RankPrompt, optimizing schema markup with InLinks, and building authority signalsβadding llms.txt is a logical next step.
2. Have Developer-Focused Content
Documentation sites, API references, and technical guides benefit most. AI developers are more likely to respect these files.
3. Want to Experiment
It takes 30-60 minutes to create and won't hurt your existing SEO. Worth testing if you're curious.
4. Think Long-Term
Even without official support today, AI companies *might* adopt this standard as AI search matures.
β Skip It If You:
1. Have Limited Resources
Focus on proven strategies first:
- Brand monitoring - Track real mentions
- Content optimization - Create cite-worthy content
- Schema markup - Structured data AI can read
- Backlink building - Build authority signals
2. Need Immediate Results
LLMs.txt is a forward-looking bet. If you need AI visibility *now*, use RankPrompt to track existing mentions and focus on getting cited more often.
3. Have a Simple Site
If your site is already clean, fast, and well-structured, AI crawlers can probably navigate it fine without llms.txt.
How to Create an LLMs.txt File (Step-by-Step)
Step 1: Audit Your Content
Identify your 20-50 most important pages:
β Product/service pages
β Best-performing blog posts (last 12 months)
β Documentation and guides
β Pricing and contact pages
β About/team information
β Old blog posts with no traffic
β Internal admin pages
β Archived content
Step 2: Create the File
Open a text editor (Notepad, VS Code, etc.) and structure your content using Markdown.
Template for Small Businesses:
```markdown
[Your Business Name]
> One-sentence pitch: What you do and who you serve
Important notes:
Services
- [Service 1](https://example.com/service-1): What it solves and who needs it
- [Service 2](https://example.com/service-2): Key benefits in 10-15 words
- [Service 3](https://example.com/service-3): Why customers choose this
Resources
- [Guide: Topic 1](https://example.com/guide-1): What readers will learn
- [Case Study: Client Name](https://example.com/case-study): Results achieved
- [Blog: Popular Post](https://example.com/blog-post): Key takeaway
Company
- [About Us](https://example.com/about): Background, team, mission
- [Contact](https://example.com/contact): How to reach us
- [Pricing](https://example.com/pricing): Plans and costs
```
Step 3: Upload to Your Website
Location: Root directory (usually `public_html/` or `/public/`)
URL Result: `https://yourdomain.com/llms.txt`
Need Help?
Step 4: Verify It Works
Visit `https://yourdomain.com/llms.txt` in your browser. You should see the plain text Markdown file.
Bonus: Run a site audit in Semrush or Ahrefs to confirm the file is being crawled.
Step 5: Maintain It
Update quarterly:
Alternatives to LLMs.txt (What Actually Works Today)
While llms.txt is experimental, these strategies have *proven* impact on AI visibility:
1. Schema Markup (Structured Data)
What: Code that tells AI systems exactly what your content represents.
Tools:
- InLinks - Automated schema generation ($49/mo)
- Schema App - Enterprise solution ($600+/mo)
- Yoast SEO - WordPress plugin (Free-$118/yr)
Why It Works: AI systems *already* use Schema.org structured data to understand content. Unlike llms.txt, this is a proven standard.
2. Brand Monitoring & Mentions
What: Track where you're mentioned online and amplify those signals.
Tools:
- Brand24 - Real-time monitoring ($99/mo)
- RankPrompt - AI-specific tracking ($49/mo)
- Onemata - LLM analytics ($249/mo)
Why It Works: AI systems learn from existing mentions across the web. More credible citations = higher likelihood of being recommended.
3. Citation-Worthy Content
What: Create content that other sites want to link to and reference.
Tools:
- Clearscope - Content optimization ($189/mo)
- Surfer SEO - AI visibility tracking ($89/mo)
- Frase - AI content research ($15/mo)
Why It Works: AI models cite authoritative sources. If your content is comprehensive and well-linked, it gets picked up.
4. Authority Signals (Backlinks)
What: Earn links from high-authority sites (DR 50+).
Tools:
- Ahrefs - Backlink analysis ($129/mo)
- Semrush - Link building outreach ($139/mo)
- Moz - Domain authority tracking ($99/mo)
Why It Works: AI systems trust content linked by established authorities. Build authority = improve AI visibility.
Concerns & Criticisms of LLMs.txt
1. Potential for "Cloaking"
The Issue: Sites could include content in llms.txt that doesn't match what users see.
Example: A business might list "We offer 24/7 support" in llms.txt while their actual site says "Email support only."
Risk: AI systems could generate misleading responses based on incorrect information.
Mitigation: Only include content that matches your visible website. Be honest.
2. No Official Adoption
Google's John Mueller compared llms.txt to the deprecated meta keywords tagβa standard that never gained traction. ([Source](https://www.sistrix.com/ask-sistrix/seo-basics/llms-txt-how-useful-is-the-file-for-llm-optimisation/))
Reality Check: We might be creating files that no one reads.
3. Maintenance Burden
Every time you add/remove content, you need to update llms.txt. For large sites, this becomes a management challenge.
Solution: Focus on evergreen content only (top 10-20 pages).
Our Experiment: LLM Relevance Directory's LLMs.txt
We implemented llms.txt on our site to test its effectiveness. View our file β
What We Included:
β Core platform pages (Tools Directory, AI Search Playbook)
β 20+ featured tools across all categories
β Educational content (comparison articles, guides)
β Vendor resources (Submit Tool, Pricing)
β Recent updates section
What We're Tracking:
- Brand mentions in ChatGPT/Claude/Perplexity (via [RankPrompt](/tools/rankprompt))
- Referral traffic from AI systems (via Google Analytics)
- Citation frequency (manual spot-checks)
Results Timeline:
We'll monitor for 6 months and update this article with findings.
Early hypothesis: Even without official AI support, creating llms.txt helped us organize our content hierarchy. Internal benefit, even if external impact is minimal.
Implementation Checklist
Before You Start:
- [ ] Audit top 20-50 pages by traffic (use [Ahrefs](/tools/ahrefs) or [Semrush](/tools/semrush))
Creating the File:
Publishing:
Ongoing:
The Bottom Line: Should You Do It?
Our Take: LLMs.txt is a low-effort hedge on the future of AI search.
Do it if:
β You have 1-2 hours to spare
β You're already optimizing for AI visibility
β You like experimenting with new standards
β Your site has complex navigation or JavaScript
Skip it if:
β You haven't done the basics (schema, monitoring, content)
β You need immediate ROI
β Your site is already clean and well-structured
The Real Priority: Focus on proven strategies firstβbrand monitoring, schema markup, citation-worthy content, and authority building. If you've done those, llms.txt is a nice-to-have addition.
Next Steps: Improve Your AI Visibility
Ready to show up in ChatGPT, Claude, and Perplexity?
Start With Proven Tools:
Track AI Mentions:
π₯ RankPrompt - See if AI already recommends you ($49/mo)
π₯ Brand24 - Monitor mentions across the web ($99/mo)
Optimize Your Content:
π₯ InLinks - Automated schema markup ($49/mo)
π₯ Clearscope - Create cite-worthy content ($189/mo)
Build Authority:
π₯ Ahrefs - Find link opportunities ($129/mo)
π₯ Semrush - Comprehensive SEO suite ($139/mo)
Learn the Fundamentals:
π [Complete AI Search Playbook (10 Steps)](/problems/show-up-in-ai-search)
π [Schema Markup for Beginners](/learn/schema-markup-simple)
π [Best AI SEO Tools for Small Business](/learn/ai-seo-tools-small-business)
Experiment Running:
We implemented llms.txt on LLM Relevance Directory. View our file at llmrelevance.com/llms.txt and check back in 6 months for results.
References:
- [Semrush: What Is LLMs.txt?](https://www.semrush.com/blog/llms-txt/)
- [NerdyData Adoption Stats](https://nerdydata.com)
- [Official LLMs.txt Specification](https://llmstxt.org)
- [Google's John Mueller on Bluesky](https://bsky.app)
- [SISTRIX Analysis](https://www.sistrix.com/ask-sistrix/seo-basics/llms-txt-how-useful-is-the-file-for-llm-optimisation/)
*Last Updated: January 14, 2025 β Implementation tested on LLM Relevance Directory*
