What is llms.txt, and does your agency need one?

Short answer: llms.txt is a plain Markdown file you put at the root of your website — youragency.com/llms.txt — listing your most important pages with a one-line description of each. It's a table of contents written for machines. Should you have one? If it costs an hour of your own time, sure. If a vendor is charging a line item for it, or calling it your ticket into ChatGPT, no — no major AI provider has confirmed using it, and Google says outright that it ignores the file.

We ship one on every site we build. We also don't pretend it's doing much. Here's the honest version, because you're going to get pitched on this file by somebody this year.

What is llms.txt, exactly?

The llms.txt proposal was published by Jeremy Howard on September 3, 2024. The problem it tries to solve is real: a language model working with a limited context window can't read your entire website, and the HTML version of your pages is mostly navigation, scripts, and styling wrapped around a few paragraphs of actual substance.

So the spec asks for a small, clean file with a fixed shape:

• An H1 with the name of the site — the only required section.
• A blockquote with a short summary containing the key information needed to understand the rest of the file.
• Zero or more H2 sections containing "file lists" — Markdown links to further detail, each with an optional note.
• An optional section named Optional, holding secondary links a model can skip when context is tight.

That's the whole standard. It is emphatically not a permissions file — it doesn't grant access, block access, or set licensing terms. It's a recommendation list.

Does anything actually read it?

This is where the pitch and the evidence part ways.

Google's guide to optimizing content for generative AI features says it plainly: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search," and "doing so will neither harm nor help your site's visibility or rankings in Google Search, as Google Search ignores them." Google added that language in June 2026 specifically because people kept asking about llms.txt. Its AI features documentation makes the same point about AI Overviews and AI Mode — no special files, no special optimizations.

Google Search Advocate John Mueller went further on Reddit, in comments widely reported in the SEO press: "AFAIK none of the AI services have said they're using LLMs.TXT (and you can tell when you look at your server logs that they don't even check for it). To me, it's comparable to the keywords meta tag." Not everyone thinks that's fair — Search Engine Land argued the opposite, since meta keywords let you claim anything while llms.txt only points at real URLs whose content still has to deliver. Decent point. Also not evidence that anyone is fetching the file.

There is one place llms.txt demonstrably earns its keep: developer documentation. AI coding assistants pull docs into context in real time, and a clean index saves them parsing a marketing site. Anthropic publishes one at platform.claude.com/docs/llms.txt, as do most modern API companies. Genuine use case. It is not your Medicare page.

Where llms.txt sits next to the files that do work

So should an insurance agent bother?

Our position, and we build these for a living: yes — but for a reason that has nothing to do with crawlers.

Writing an llms.txt forces you to answer, in about eight sentences, questions most agency websites never answer anywhere: what you sell, which states you're licensed in, which counties you actually serve, what it costs to work with you, and what someone should do next. Almost every agent we talk to has those answers in their head and nowhere on their site. The exercise is worth more than the file.

The second reason is optionality. If OpenAI, Anthropic, or Perplexity ever announce they read it, you're already done; if they never do, you spent an hour. What you should not do is pay a setup fee for it, accept it as the deliverable in an "AI optimization" package, or believe anyone who says it produces citations. Nobody can promise that, from this file or any other.

Ours lives at foundagents.ai/llms.txt — open it and copy the shape.

What to put in yours, if you write one

1. H1 with your agency's legal name. The name on your license, matching your Google Business Profile and your website footer exactly.
2. A blockquote summary. Lines of business, states you're licensed in, counties you serve, how someone reaches you. Write it so a stranger quoting only that paragraph would still describe your agency correctly.
3. A "Key answers" section. Direct question-and-answer pairs: how you're compensated, whether you're independent or captive, how fast you respond. It doubles as source material for your FAQ pages.
4. Your main pages, as Markdown links with a one-line description each. Home, about, each line of business, contact.
5. Your county or city pages, in their own section, named by county.
6. Recent articles. Title, URL, and one line on what question each one answers.

Two rules. Every URL must resolve — a file pointing at deleted pages is worse than no file. And nothing goes in it that isn't also visible on your site: if you sell Medicare, your consumer-facing pages still need the "we do not offer every plan available in your area" language and the non-affiliation disclaimer, and a text file is no place for claims you wouldn't put on the page. Confirm specifics with your FMO and carriers.

If you only have one afternoon, do these instead

1. Open youragency.com/robots.txt and confirm you aren't blocking AI crawlers. This is the single highest-consequence line of text on your website.
2. Fix your schema markup. InsuranceAgency on the homepage, Person for the licensed agent, areaServed naming real counties. Our page-by-page schema map has the full list.
3. Rewrite your three most important pages answer-first — the direct answer in the first two sentences, then the detail. That's the shape a model can quote without a paraphrase.
4. Publish something only you can write. Google's own guidance says unique, useful content will influence your presence in generative AI search more than anything else in its guide. Your carrier lineup by county, what you tell clients about a specific plan change, what you actually charge.

That's the real order of operations, and it's the same one in our AEO guide for insurance agents. llms.txt is item eleven on a ten-item list — worth doing, never worth doing first. Every site we ship gets one anyway, along with the nine things above it; the full deliverable list is on what you get.

Frequently asked questions

What is an llms.txt file?

It is a plain Markdown file placed at the root of a website — yoursite.com/llms.txt — that gives AI systems a curated index of the site's most important pages. The proposal, published by Jeremy Howard in September 2024, calls for an H1 with the site name, a blockquote summarizing what the site is, and H2 sections listing links with a one-line description each. It is a table of contents, not an access-control file: it neither grants nor blocks any crawler.

Does Google use llms.txt?

No. Google's guide to optimizing content for its generative AI features states that you don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search, and that doing so will neither harm nor help your site's visibility or rankings in Google Search, as Google Search ignores them. Google added that clarification in June 2026 in response to community questions. Publishing one is allowed and, as far as Google Search is concerned, has no effect either way.

Is llms.txt the same as robots.txt?

No, and confusing the two is the expensive mistake. robots.txt is an access-control file that crawlers have honored since 1994: it tells GPTBot, ClaudeBot, PerplexityBot and everyone else which URLs they may fetch, and blocking them there makes you invisible to those assistants. llms.txt is a suggestion — a curated list of pages you would like a model to read if it happens to look. If your robots.txt blocks AI crawlers, no llms.txt file will undo it.

Will an llms.txt file hurt my SEO?

No. Google has said explicitly that maintaining llms.txt or similar files will neither harm nor help visibility or rankings in Google Search, because Google Search ignores them. The only real cost is opportunity cost. An outdated llms.txt that lists pages you deleted is worse than not having one, and an hour spent on the file is an hour not spent on crawler access, schema markup, and content actually worth quoting.

How do I create an llms.txt file for my agency website?

Open a text editor. Write an H1 with your agency name, then a one-paragraph blockquote covering what you sell, which states you are licensed in, and which counties you serve. Below that, add H2 sections listing your main pages and county pages as Markdown links with a one-line description each. Save it as llms.txt and upload it to the root of your domain so it loads at youragency.com/llms.txt. It takes under an hour, needs no plugin, and nobody should charge you a setup fee for it.