LLMs.txt Generator

This guide explains the llms.txt standard from the ground up: what an llms.txt file is, what it is good for, what the markdown format actually looks like, how it differs from robots.txt, why it matters for AI visibility, and how to create and maintain one over time. llms.txt is a proposed, voluntary convention (introduced by Jeremy Howard of Answer.AI in September 2024) for a simple markdown file placed at your website root that hands large language models a curated, link-first map of your most important content. If your goal is to be understood and cited by AI assistants like ChatGPT, Claude, Perplexity, and Gemini, this file is one of the cleanest signals you can offer them. And once it is live, you can use the free AmICited AI Brand Visibility Report to check whether those assistants actually mention and cite your brand.

We will keep this educational and accurate. llms.txt is a community proposal, not a ratified web standard, and adoption is optional, so we will be careful not to promise rankings or invent numbers. What we can say is that the format is easy to publish, costs you almost nothing, and aligns your site with how AI systems prefer to read the web.

What is llms.txt?

llms.txt is a plain markdown file that lives at the root of your domain, at yoursite.com/llms.txt. Its job is to give large language models a concise, human-readable and machine-readable summary of your site, plus a curated list of links to the pages that matter most. Think of it as a guided tour rather than a locked door. Instead of forcing an AI system to crawl, render, and parse dozens of HTML pages full of navigation, scripts, and ads, you offer it a single flattened document that says, in effect, here is who we are and here is what is worth reading.

Jeremy Howard, co-founder of Answer.AI, first proposed the idea in September 2024. The motivation is practical: web pages are built for people, so they are noisy and expensive for models to process, and context windows are limited. A curated markdown file in a predictable location conserves that context and reduces guesswork. It is worth being precise about status, though. llms.txt is a proposed, informal standard. Neither the IETF nor the W3C has ratified it, and support from AI vendors is voluntary and still evolving.

What is llms.txt good for?

The file solves a specific problem: helping AI systems find and understand your best content quickly and accurately. When an assistant can read a clean summary and follow labeled links, it is more likely to represent your site correctly rather than stitching together a fuzzy impression from scattered pages.

Curating the pages you most want AI systems to reference, such as documentation, product pages, pricing, and cornerstone guides.
Giving each link a short description so a model understands the purpose of a page before it fetches it.
Reducing ambiguity about what your company does, which lowers the chance of an assistant describing you incorrectly.
Saving model context and processing effort by replacing messy HTML with lean markdown.
Providing a single, stable place to signal your priorities as your site grows and changes.

What an llms.txt file looks like

The format is deliberately minimal. It opens with an H1 that names the site or project, followed by a blockquote that summarizes what you do. After that you can add optional prose, then one or more H2 sections, each holding a markdown bullet list of links. Every link can carry a short note after a colon so the model knows what it will find. A common convention is a final section titled Optional, which lists secondary pages an assistant can skip when its context budget is tight.

Here is a short, realistic example for a fictional analytics company:

# Northwind Analytics

> Northwind Analytics is a self-serve product analytics platform that helps SaaS teams track activation, retention, and revenue in one place.

Use the links below to understand our product, pricing, and documentation. All linked pages are available as clean markdown.

## Core pages

- [Product overview](https://northwind.example/product): What Northwind does and who it is for
- [Pricing](https://northwind.example/pricing): Plans, limits, and what each tier includes
- [Getting started guide](https://northwind.example/docs/start): Install the SDK and send your first event

## Documentation

- [API reference](https://northwind.example/docs/api): Endpoints, authentication, and rate limits
- [Data model](https://northwind.example/docs/model): How events, users, and sessions relate

## Optional

- [Company blog](https://northwind.example/blog): Product updates and analytics tutorials
- [Changelog](https://northwind.example/changelog): Recent releases and fixes

llms.txt vs robots.txt

These two files are often confused, but they do opposite jobs. robots.txt is a permission and access file. It tells crawlers which URLs they may or may not fetch, using directives like User-agent, Allow, Disallow, and Sitemap. It contains no content from your site, only rules, and it has been a recognized web convention for decades.

llms.txt is a content and context file. It does not grant or deny access to anything. Instead it recommends what is worth reading and explains what each page is about, using structured markdown with headings, a blockquote, and described links. One controls the door; the other draws the map. They are complementary, and many sites publish both. If you also want to manage how AI crawlers specifically are allowed to access your site, that is a robots.txt job, and you can assemble those rules quickly with the free AmICited robots.txt for AI generator.

How it helps your business and why it matters for AI visibility

Search is shifting from a list of blue links toward direct answers generated by AI assistants. In that world, being cited inside an answer is the new version of ranking on page one. This is the heart of generative engine optimization, or GEO: shaping how AI systems perceive, summarize, and reference your brand.

An llms.txt file supports that goal by making your best content easy to locate and hard to misread. When an assistant can quickly grasp what you offer and follow clearly labeled links, it is better positioned to mention you accurately and point users to the right page. It will not guarantee citations, and no honest tool can promise that, because adoption is voluntary and vendor behavior varies. What it does is remove friction and reduce the chance of being misrepresented, which is a reasonable and low-cost bet.

The practical next step is measurement. Publishing the file is an input; visibility is the outcome you actually care about. The free AmICited AI Brand Visibility Report shows whether assistants surface and cite your brand for the queries that matter to you, and the AmICited ChatGPT query fan-out generator helps you map the many related questions a single user prompt can expand into, so you can see where you already appear and where you are missing.

How to create and maintain an llms.txt file

Creating the file is straightforward, and maintaining it is mostly a matter of keeping it honest as your site evolves.

List your most important pages first: what you do, your core products, pricing, and your best documentation or guides.
Write a tight blockquote summary in plain language, one or two sentences that a model could quote directly.
Give every link a short description after a colon so its purpose is unmistakable.
Group links under clear H2 headings, and move nice-to-have pages into an Optional section.
Save it as markdown named llms.txt and place it at your domain root so it resolves at yoursite.com/llms.txt.
Consider a companion llms-full.txt, an optional convention that packs full page content into one long markdown document for models that can ingest more depth.
Revisit it on a schedule. When you launch products, restructure docs, or change pricing, update the links and the summary so the file never drifts from reality.

Frequently asked questions

Is llms.txt an official web standard?

No. llms.txt is a proposed, informal convention introduced by Jeremy Howard of Answer.AI in September 2024. It has not been ratified by the IETF or the W3C, and support from AI vendors is voluntary. Adoption has grown, and some tooling now checks for the file, but you should treat it as a helpful, low-cost signal rather than a guaranteed requirement. Publishing one is optional, and it will not by itself force any assistant to read or cite your site.

Where should the llms.txt file be placed?

Put it at the root of your domain so it resolves at yoursite.com/llms.txt, the same pattern used by robots.txt. The predictable location is part of the point, because it lets any AI system look in one known place. Save it as plain markdown with the exact filename llms.txt. If you run multiple subdomains or distinct properties, each can host its own file at its own root, tailored to the content that lives there.

What is the difference between llms.txt and llms-full.txt?

llms.txt is the concise map: an H1 title, a blockquote summary, and curated sections of described links. llms-full.txt is an optional companion convention that packs the full content of your key pages into one long markdown document, so a model that can handle more depth does not need to fetch many separate pages. Think of llms.txt as the table of contents and llms-full.txt as the bundled book. The full version is heavier to maintain, so many sites start with llms.txt alone.

Does llms.txt replace robots.txt or my sitemap?

No, they serve different purposes and can coexist. robots.txt controls crawler access with allow and disallow rules, and an XML sitemap lists URLs to help search engines discover pages. llms.txt does neither of those things. It curates and describes your most valuable content for large language models, adding context that the other files never carry. Keep robots.txt for access control, keep your sitemap for discovery, and add llms.txt to shape how AI systems understand what matters.

Will publishing llms.txt guarantee that AI assistants cite my site?

No, and any tool that promises guaranteed citations is overstating things. Because adoption is voluntary and vendor behavior differs, the file cannot force an assistant to read it or reference you. What it does is make your best content easy to find and hard to misread, which reduces friction and the risk of misrepresentation. To see real outcomes, measure them. The free AmICited AI Brand Visibility Report shows whether assistants actually surface and cite your brand.

How often should I update my llms.txt file?

Update it whenever your site changes in ways that affect what you want AI systems to know: new products, restructured documentation, changed pricing, or retired pages. A good rhythm is to review it alongside major releases and then do a lighter quarterly check. The main risk is drift, where the file describes a version of your site that no longer exists. Keeping the summary and links honest is more valuable than adding volume, so prune stale entries as readily as you add new ones.