Perplexity Adds llms.txt Support

What llms.txt support means for Perplexity

When Perplexity processes a query, it now checks for anllms.txt file at the root of your domain. If present, the model uses it as guidance for:

• Content hierarchy — Which pages to prioritize for retrieval
• Entity context — Descriptions of key entities and topics
• Usage guidance — How you'd prefer your content to be cited
• Freshness signals — Which pages are recently updated

This is a significant departure from passive crawling — Perplexity is now actively consulting your llms.txt as part of its answer generation pipeline.

Writing an effective llms.txt for Perplexity

Perplexity's implementation favors structured, clear guidance. Here's what works:

1. Start with a summary. One paragraph describing what your site covers and your target audience.
2. List your key sections. Bullet points or numbered sections with descriptions help Perplexity navigate your content.
3. Name your key entities. List your company, products, team members, and key topics with brief descriptions.
4. Link to your best pages. Direct Perplexity to your most authoritative, well-structured content.
5. Don't block crawlers. llms.txt is guidance, not a directive. Ensure PerplexityBot can actually access the pages you list.

Beyond Perplexity: preparing for other AI engines

While ChatGPT doesn't formally support llms.txt directives yet, Anthropic has acknowledged it in ClaudeBot documentation. As more AI engines adopt this standard, implementing llms.txt becomes a cross-platform strategy for content prioritization.

A well-structured llms.txt today prepares you for tomorrow's AI search landscape — where content discoverability and AI citation frequency are increasingly tied together.