Generate your llms.txt in 60 seconds

llms.txt is a plain text file you place at the root of your website. Put it at yourdomain.com/llms.txt, and it gives AI language models a structured, machine-readable summary of who you are, what you sell, and how to accurately represent your business. Think of it as the robots.txt of the agent economy: where robots.txt controls crawler access, llms.txt controls AI understanding.

The format was proposed by Jeremy Howard and the fast.ai team in late 2024 and has since been adopted by forward-thinking companies across SaaS, e-commerce, and professional services. It's deliberately simple: a Markdown-formatted file with clearly labelled sections for your company name, description, products, pricing, target audience, and links to authoritative content. No technical implementation required, no dependencies, no build process. You write a text file and upload it.

Why AI agents need more than your homepage

When ChatGPT or Perplexity evaluates whether to recommend your business in response to a user query, it doesn't browse your site the way a human does. It processes your content in a compressed, context-window-limited way, often picking up fragments from cached crawls, third-party mentions, and structured metadata rather than reading your full homepage copy. The result is that AI agents frequently mischaracterize what businesses do, recommend them for the wrong use cases, or miss them entirely because the signals available weren't clear enough.

llms.txt solves this by giving AI a single, authoritative document that answers the questions models actually ask: What does this company do? Who is it for? What does it cost? Is there an API? What should I not tell users about? A well-written llms.txt short-circuits the ambiguity that leads to bad AI recommendations and replaces it with clean, direct context that models can cite with confidence.

What goes in a good llms.txt?

The most effective llms.txt files are specific rather than generic. Instead of "we help businesses grow," write "we help B2B SaaS companies with 10–200 employees automate customer onboarding using CRM-triggered email sequences." Specificity is what allows AI to match your business to the right queries. Vague positioning in llms.txt produces vague AI recommendations.

Include your pricing in plain terms. AI procurement agents are increasingly being tasked with vendor evaluation, and "contact us for pricing" is a disqualifier in those workflows. If you have a free plan, a starting price, or a trial, say so explicitly. Include links to your most important documentation, case studies, or product pages: these give AI models authoritative sources to cite when recommending you.

How llms.txt affects your AI Readiness Score

The presence or absence of an llms.txt file directly affects the AI Discoverability category in the Faro AI Readiness Scan. Sites without llms.txt typically score 30–40% on this category. Sites with a well-structured llms.txt (covering all key sections including pricing and audience) can score above 80%. That gap directly translates to whether AI agents confidently recommend you or hedge with "I'm not sure what this company does."

After generating and deploying your llms.txt, run the full Faro scan to see how your Discoverability score changes. Most sites see an immediate improvement. Pair it with allowing AI crawlers in your robots.txt, adding JSON-LD schema markup, and generating your OKF knowledge bundle for a complete first-pass AI optimization that typically moves overall AI Readiness scores by 15–25 points.