Short answer: an llms.txt file is a Markdown file at the root of your site (yoursite.com/llms.txt) that gives AI tools a short, curated index of your site: a one-line summary of what the site is, then your important pages grouped under headings, each with a one-line description. Its companion llms-full.txt holds the full text of those pages. You can write them by hand, or generate both from your sitemap with the llms.txt Generator, which costs $0.004 per page, so a 50-page site costs 20 cents.
llms.txt is one of the most talked-about ideas in AI search optimisation and one of the most over-sold. This guide explains what it is, gives an honest answer about who uses it, shows the format, and walks through generating one. This site has served its own llms.txt and llms-full.txt for months, so the advice comes from running it, not from theory.
What is llms.txt?
llms.txt is a proposed standard, published at llmstxt.org, for a Markdown file at a website's root that helps large language models understand the site. The idea is simple: HTML pages are full of navigation, scripts and layout that waste an AI model's limited context. A plain Markdown index that says "this is what we are, and these are the pages that matter" is far easier for a model or an AI agent to use.
- llms.txt is the index: a title, a short summary, and lists of links with one-line descriptions, grouped into sections.
- llms-full.txt is the full version: the clean Markdown content of every listed page in one file, so an AI tool can read the whole site in a single request.
Does llms.txt actually help with ChatGPT, Perplexity or Google?
An honest answer, because many articles skip it: llms.txt is not a ranking factor, and Google has said its search does not use it. Adding the file will not move your Google rankings. The major AI companies have not committed to reading it as part of how their assistants answer questions either.
So why bother? Three reasons that hold up:
- AI agents and developer tools do read it. Coding assistants, documentation tools and agents that browse on a user's behalf increasingly look for llms.txt, because it is the cheapest way to understand a site. It matters most for documentation, APIs and product catalogues.
- It forces you to state clearly what you are. Writing one sentence that describes your business, and choosing the 20 pages that matter, is useful positioning work that also improves your homepage and meta descriptions.
- It costs almost nothing. Generating it takes minutes and a few cents, and keeping it current can be scheduled.
Treat it as cheap hygiene for AI readability, not a growth hack. The things that actually get you cited by AI answer engines are clear answers on your pages, structured data, and being mentioned by sources those engines trust.
What should an llms.txt file look like?
The format is plain Markdown with a fixed structure: an H1 with the site name, a blockquote summary, optional notes, then H2 sections containing link lists.
# Acme Analytics
> Acme Analytics is a privacy-first web analytics platform for small teams.
## Main Pages
- [Acme Analytics](https://acme.io): Privacy-first analytics for small teams.
- [Pricing](https://acme.io/pricing): Simple plans starting free.
## Docs
- [Getting Started](https://acme.io/docs/getting-started): Install the snippet in five minutes.
Rules that keep it useful:
- Make the summary specific. "We help businesses grow" says nothing. Name what you sell, to whom, and where.
- Curate. Pricing, product or service pages, documentation and your best guides. Leave out tag pages, archives, legal boilerplate and thin pages.
- Write real descriptions. One line per link that says what the page answers.
- Keep it current. A stale llms.txt that lists deleted pages or old prices is worse than none.
A good addition, and one many generated files miss, is a short guidance section in your own words: when your site is the right source to cite and when it is not. This site's llms.txt does exactly that.
How do you generate llms.txt automatically?
- Open the llms.txt Generator and paste your homepage URL.
- Set Max pages. It reads your sitemap first; with no sitemap it crawls internal links.
- Focus it with Include URL patterns such as
/docs/or/blog/, and keep noise out with Exclude URL patterns such as/tag/or/cart. - Leave "Generate llms-full.txt" on if you want the full-content file as well.
- Run it, then download both files from the run's key-value store. The dataset lists every page with its title, description, section and word count, so you can check what was included.
- Edit the summary line by hand, since that one sentence deserves a human, then upload both files to your site root.
- Schedule it weekly so the file follows your content.
How much does it cost?
$0.004 per page processed. A 50-page site costs 20 cents; a 500-page site costs $2. On the free Apify plan you can process up to 10 pages per run, enough for a small site. A paid Apify plan allows up to 500 pages per run and scheduling. You can generate one on Apify now.
How do you upload llms.txt?
- Static sites and most hosts: put the file in the public root folder, next to robots.txt.
- WordPress: upload it to the site root by FTP or your host's file manager, or use a plugin that serves it.
- Shopify and other hosted platforms: many do not allow files at the root; use a redirect from
/llms.txtto a file you can host, or a platform app.
Check it by opening yoursite.com/llms.txt in a browser: it should display as plain text.
What else makes a site readable for AI?
llms.txt is one small piece. The bigger ones are pages that answer questions directly in the first paragraph, FAQ sections with matching structured data, clear entity information about who you are, and fast, crawlable HTML. The SEO, GEO and AEO audit checks a page for all of these, and the AI Brand Visibility Tracker tells you whether AI assistants actually mention you. For the bigger picture, see is ChatGPT recommending your competitors.
Frequently asked questions
A Markdown file at a website's root that gives AI models a curated index of the site: a title, a one-line summary and grouped links with short descriptions. It is a proposed standard from llmstxt.org.
llms.txt is a short index of links with descriptions. llms-full.txt contains the full Markdown text of the listed pages in one file, so an AI tool can read the whole site at once.
No. It is not a Google ranking factor and Google has said its search does not use it. Its value is making your site easier for AI agents and tools to read, at almost no cost.
Use a generator that reads your sitemap, such as the llms.txt Generator on Apify. Paste your URL, run it, download llms.txt and llms-full.txt, edit the summary line and upload both files to your site root.
$0.004 per page with the llms.txt Generator, so a 50-page site costs 20 cents. The free Apify plan covers up to 10 pages per run.
See all the tools for search and AI visibility in the SEO and AI visibility hub.
β‘ Run this without building it
These actors already do what this guide describes. Free $5 credit on a new Apify account.
π Keep reading
Get the Free Web Scraping Toolkit
Join the newsletter and get my curated list of scraping tools, proxy comparison cheatsheet, and Python automation templates.