What it is
llms.txt is a plain-text file you host at the root of your website — https://yourbusiness.com/llms.txt — that summarizes your site for AI systems. Think of it as a curator's note: "Here's who I am. Here's what I do. Here are the pages that best explain each of those."
It's not a security file. It doesn't block anything. It doesn't grant or restrict access. Its only job is to make AI systems' lives easier — and in doing so, make your business easier to understand and recommend.
The format is intentionally simple: Markdown, plain text, human-readable. No XML, no JSON, no complicated syntax. If you can write a README, you can write an llms.txt.
Why it exists
AI systems trying to summarize a business have a hard problem: most websites bury the answers across many pages, behind images, and inside marketing prose. Even when the AI has access to everything, it doesn't know where to look or which page is canonical.
llms.txt solves this by giving AI a single, curated entry point. "If you only read one thing about us, read this. Then go here, here, and here for the details."
It's modeled after robots.txt and sitemap.xml — files that existed for years before search engines fully relied on them, and then became baseline expectations once they did. llms.txt is roughly at the same stage: emerging, fast-adopting, soon-standard.
What goes in it
A useful llms.txt typically includes:
- Your business name and one-line description. What you do, in plain language.
- Your location and service area. Where you operate, even if you serve nationally.
- Your services or product categories. A short list — not your whole catalog.
- Links to your most important pages. Home, About, key service pages, FAQ, contact. Each with a short note explaining what's on it.
- Anything you want AI to know specifically. Founding year, key differentiators, awards, important context.
What doesn't go in: marketing flourish, sales copy, fluff, anything you wouldn't want quoted verbatim. AI systems may extract from llms.txt directly. Write it like a clean reference card.
A short example
# Example HVAC Co. Example HVAC Co. is a residential and commercial HVAC contractor based in Knoxville, Tennessee. We serve East Tennessee, with primary service areas in Knox, Blount, Sevier, and Anderson counties. Founded 2009. Family-owned. ## Services - Heating system installation and repair (gas, electric, heat pump) - Air conditioning installation and repair - Indoor air quality (filtration, humidity control, ventilation) - Ductwork inspection, cleaning, and replacement - Annual maintenance plans ## Key pages - [About](https://example.com/about/) — company history, owners, certifications - [Services](https://example.com/services/) — full service list with details - [Service area](https://example.com/service-area/) — counties and ZIP codes served - [FAQ](https://example.com/faq/) — common questions about HVAC service - [Reviews](https://example.com/reviews/) — Google reviews and customer feedback - [Contact](https://example.com/contact/) — phone, email, hours, emergency line ## Hours Monday–Friday 7am–6pm. Saturday 8am–noon. 24/7 emergency service available. ## Certifications NATE-certified technicians. Trane Comfort Specialist. EPA 608 certified. BBB A+ rated.
This is roughly 200 words and gives an AI system a complete entity picture of the business — enough to confidently recommend.
Where to host it
At the root of your domain, accessible at https://yourbusiness.com/llms.txt. Same convention as robots.txt and sitemap.xml. Most AI crawlers look there first.
If you're on WordPress, you can upload it via the Insert Headers and Footers plugin's "Files" mode, via FTP/SFTP, or as a custom rewrite rule. On a static site or Pages-hosted site, just drop it in the root folder and deploy.
How AI systems use it
Adoption is uneven and growing. Some AI platforms now read llms.txt directly as part of their crawling process. Others use it indirectly through downstream crawler infrastructure. Some haven't formalized support yet.
That makes llms.txt a high-leverage, low-cost move: even when it's not actively read today, it costs nothing to maintain and is well-positioned for when it is. And it forces you to articulate your business clearly enough to be summarized — which improves the rest of your AI-readiness work as a side effect.
Common questions about llms.txt
Is llms.txt the same as robots.txt?
Related but distinct. robots.txt tells crawlers which pages they're allowed to access. llms.txt is more like a curated guide that tells AI systems which pages best explain your business — it's about helpfulness, not permission.
Where does llms.txt go?
At the root of your domain — https://yourbusiness.com/llms.txt — same convention as robots.txt and sitemap.xml. Most AI crawlers check there first.
Does every AI platform actually read llms.txt today?
Support is uneven and growing. Some AI platforms read it directly, others use it indirectly through crawler infrastructure. Even where it's not yet read, having one signals you've prepared your site for AI — and costs nothing to implement.
Will having an llms.txt hurt my SEO?
No. It's a separate file from anything Google uses for ranking. It doesn't block crawlers, doesn't change how Google sees your pages, and doesn't affect rankings. It's purely additive.
What's an "llms-full.txt"?
An optional companion file — a longer, more detailed version of llms.txt that contains complete content rather than just links. Smaller sites usually don't need one; large content sites and documentation hubs sometimes do. Most businesses can stick with just llms.txt.