llms.txt is a proposed plain-text file you place at the root of your site to hand AI models a clean, curated map of your most important pages. It is optional, it is not an official standard, and right now it does almost nothing for how ChatGPT, Gemini, or Google actually see you. Here is the honest version of what it does, what it doesn’t, and when it’s worth ten minutes of your time.

What llms.txt actually is

llms.txt is a single Markdown-formatted text file that lives at your domain root — yoursite.com/llms.txt. The idea, first proposed in late 2024, is simple: modern web pages are buried under navigation, ads, cookie banners, and scripts, and large language models waste effort digging your real content out of that clutter. So you write a stripped-down index — your site title, a one-line description, and a curated list of your best, most current pages — and point AI systems straight at it.

Think of it as an executive summary written for machines, not a sitemap. A typical file looks like this in plain terms:

• An H1 title — your brand or site name.
• A one-sentence summary — what you do, in a single blockquote line.
• Grouped links under H2 headings — “Services,” “Guides,” “Documentation,” each with a handful of your canonical URLs and a short note on what each covers.

There is also a heavier variant, llms-full.txt, which dumps the actual text of your key pages into one long file so a model can ingest everything in a single pass. That one is mostly used by software companies publishing developer documentation, not by service businesses.

To make it concrete, a Nevada plumbing company’s file might open with an H1 of the business name, a one-line summary such as “Licensed plumbing and water-heater service across the Las Vegas valley,” then an H2 for “Services” listing the repair, installation, and emergency pages, and an H2 for “Guides” pointing at its most useful blog posts. That’s the whole file — short, honest, and readable in a single glance. The discipline it forces — deciding which handful of pages actually matter — is arguably more useful than the file itself.

Rule of thumb: llms.txt is a suggestion you leave for AI crawlers, not a rule they’re obligated to follow. Treat it as a nice-to-have, never as the thing that makes or breaks your visibility.

llms.txt vs robots.txt (and does it actually work)

People reach for the robots.txt comparison because both files sit at the root and both are aimed at crawlers. But they do opposite jobs, and only one of them is actually respected across the board.

robots.txt tells crawlers where they may NOT go

robots.txt is a decades-old, near-universally-honored convention. It’s a gatekeeper — it disallows bots from crawling certain paths. Googlebot, Bingbot, and the major AI crawlers all check it and generally obey it. It’s part of the plumbing of a healthy site, which is why it lives in the world of technical SEO alongside your XML sitemap, canonical tags, and crawl budget.

llms.txt tries to tell models where they SHOULD look

llms.txt is the inverse — a curated invitation rather than a restriction. And unlike robots.txt, it has no enforcement and no broad adoption. That’s the part most articles skip over, so let’s be blunt about the evidence:

• Almost nobody’s bots read it. Large-scale crawl studies have found the overwhelming majority of llms.txt files receive zero bot requests. The file can exist and simply never get fetched.
• Google has said it doesn’t use it. Google’s own search team has stated plainly that files like llms.txt are not a ranking or citation signal for Search or AI Overviews. Creating one won’t help or hurt your rankings.
• It’s unverified, so AI discounts it. Because the site owner controls the file completely, it’s easy to stuff or game — the same weakness that killed the old meta keywords tag. AI systems lean on what they can verify against your live HTML, not a self-declared cheat sheet.

So the honest answer to “does it work” is: not in any measurable way today, for the general web. The developer-documentation world (think API references and product docs) is the one place it sees real use, because a few coding assistants do fetch it there.

Should your site have one?

For most Nevada businesses, adding llms.txt is a low-cost, low-reward move. It won’t hurt you, and if the standard gains traction later you’ll already be set up. But it should sit at the very bottom of your priority list, well below the fundamentals that genuinely influence how you show up in AI answers. Here’s how we’d decide:

1. Add one if you publish documentation or a large reference library. Software, SaaS, and technical-product sites get the most out of it because AI coding tools actually consume it. If that’s you, it’s worth the ten minutes.
2. Skip it (for now) if you’re a local service business. A plumber, law firm, or medical practice in Las Vegas or Reno gains nothing measurable today. Your time is better spent on the pages, schema, and reviews that AI does read.
3. If you add it, keep it honest and current. List only your canonical, high-value URLs, keep the descriptions accurate, and update it when your site changes. A stale llms.txt is worse than none.
4. Never treat it as a substitute for real optimization. The heavy lifting for AI search happens in your content and markup, which is the domain of answer engine optimization — structuring pages so they directly answer the questions people ask.

If you do decide to build one, it takes about ten minutes: open a plain-text editor, write your title and one-line summary, group your canonical URLs under a few clear headings, save the file as llms.txt, and upload it to your site’s root so it resolves at yoursite.com/llms.txt. Free generators will scaffold one from your sitemap, but review the output — they tend to dump in every page, and curation is the entire point.

Our take: it’s an interesting proposal worth watching, not a task worth stressing over. We add it for clients where it makes sense and quietly skip it where it doesn’t. Nothing about your 2026 visibility hinges on it.

Frequently asked questions