Call · 30 min
GlossaryMattia Esposito10 November 20264 min read

llms.txt. The file that summarises a site for AI assistants, and who says they read it.

llms.txt is a Markdown file, proposed in 2024, that summarises a site for language models. Here is how it is built, how it differs from robots.txt and what Google says about it.

In brief

llms.txt is a Markdown file at the root of a site that summarises it for language models: a title, a summary and a list of useful pages.

A proposal, not a standard, published on llmstxt.org in September 2024. The only required part is the title.

Google Search ignores it, and the crawlers of OpenAI, Anthropic and Perplexity don’t state that they use it to choose sources.

Written by: Mattia Esposito, AI and marketing consultant, founder of Itria AI. From Puglia, based in Turin since 2023, he works remotely in Italy and abroad. LinkedIn profile.

How an llms.txt is built

An llms.txt is a Markdown file with one required title and a few optional parts. The proposal, described on llmstxt.org and updated to its second version in August 2026, puts it this way: “An H1 with the name of the project or site. This is the only required section”.

PartContents
# Title

The name of the site. The only required part.

> Summary

A blockquote with the key information in a few lines.

Notes

Paragraphs or lists without headings, with useful details.

## Sections

Lists of links to pages, each with an optional note.

## Optional

By convention, secondary material an agent can skip.

An annotated example, line by line, is in the guide to llms.txt; how to add one to a WordPress or Shopify site, with the plugins and settings, in the guide to llms.txt for WordPress and Shopify.

llms.txt, robots.txt and sitemaps

All three files sit at the root of the site and do three different jobs. robots.txt grants or denies permission to read, and it is a standard, RFC 9309, published by the IETF in September 2022. The sitemap lists the addresses to discover. llms.txt summarises the site for models and grants no permissions: to keep an assistant out you use robots.txt, as the guide on robots.txt and AI crawlers explains.

Who reads it

Not Google Search. Its guide to AI features in Search, updated on 10 July 2026, says creating one neither helps nor hurts: “Google Search ignores them”. OpenAI, Anthropic and Perplexity, in the documentation on their crawlers read on 10 November 2026, don’t state that they use it to choose sources.

On 7 September 2026 I published my analysis of the websites of Italian companies that also sell abroad: 28 out of 143 had an llms.txt, and 24 of those had been written by a plugin.

In August I had written my own llms.txt, along with the structured data, and ChatGPT still didn’t know who had founded itria.io. An assistant cites a page for other reasons: doors open to AI crawlers and content that can be checked, the work this glossary calls GEO.

Questions and answers

What is the llms.txt file?

A plain-text Markdown file placed at the root of the site, at /llms.txt, that summarises the site for language models: a title, a short description and a list of useful pages with a link to each.

Does llms.txt help you appear on Google?

No. In its May 2026 guide to AI features, Google writes that Search ignores it: creating one neither helps nor hurts. You can keep it for other services that use it.

What is the difference between llms.txt and robots.txt?

robots.txt tells crawlers what they may read, and it is an IETF standard, RFC 9309. llms.txt is a proposal that summarises the site for models, and it neither grants nor removes permissions.

What should an llms.txt contain?

The only required element is an H1 title with the name of the site. Then, according to the proposal, a summary in a blockquote, any notes, and H2 sections with lists of links to pages.

How many sites have an llms.txt?

In Mattia Esposito’s analysis of the websites of Italian companies that sell abroad, 28 out of 143 had the file, and 24 of those had been written by a plugin.

Notes on sources

  1. llmstxt.org, The /llms.txt file, version 2 (August 2026), and changes page: the quotation. Read on 10 November 2026.
  2. Google Search Central, Optimizing your website for generative AI features on Google Search, updated 10 July 2026, CC BY 4.0.
  3. IETF, RFC 9309, Robots Exclusion Protocol, September 2022.
  4. The figure on 143 sites comes from Mattia Esposito’s analysis published on LinkedIn on 7 September 2026.
·Who does it for you

Writing the file takes ten minutes. Getting chosen by assistants instead of your competitors is the work that follows. Positioning is built starting from the market.

Mattia Esposito, who wrote this guide, also works with Itria on positioning: in-depth market analysis to find where competitors are exposed and take that ground. He did it on his own site first. In August, asked who had founded itria.io, ChatGPT couldn’t find him. From 11 October to 7 November 2026 itria.io pages appeared 13,803 times in Google and Bing results and 3,818 times in AI answers: Google’s AI Overviews and AI Mode, and Microsoft Copilot.