Signals that matter

llms.txt and schema markup: what actually matters for AI search

By Mo Buhariwala, founder of Namova Last updated

llms.txt is a proposed text file that summarises a website for AI tools, and schema markup is code that labels facts on a page for machines. Schema is an established standard that Google reads and recommends; llms.txt is a proposal that no major AI company has said it relies on for answers. For most businesses, the technical work that matters more is simpler: letting the AI search bots read your site and keeping your facts accurate.

This guide explains both, what the official documentation says, and a short checklist of what to do.

What is llms.txt?

llms.txt is a plain text file placed at the root of a website (for example, yoursite.com/llms.txt) that gives AI tools a short summary of the site and links to its most useful pages. It was proposed by Jeremy Howard in September 2024. The official site describes it as "a proposal to standardise on using an /llms.txt file to provide information to help agents use a website" (llmstxt.org).

The word that matters is "proposal". llms.txt is not an agreed web standard like robots.txt, and the AI companies have not committed to reading it.

Does llms.txt work?

There is no public evidence that llms.txt makes ChatGPT, Claude, Gemini or Perplexity name a business more often. Here is what each company's own documentation focuses on:

CompanyWhat its documentation tells site owners to manageMentions llms.txt?
GoogleNormal crawling and indexing; "You don't need to create new machine readable files, AI text files, or markup to appear in these features"No, and says no AI text files are needed
OpenAIrobots.txt rules for OAI-SearchBot (search) and GPTBot (training)No
Anthropicrobots.txt rules for Claude-SearchBot, Claude-User and ClaudeBotNo
Perplexityrobots.txt rules for PerplexityBot; Perplexity-User, which fetches pages a person asks for, "generally ignores robots.txt rules"No

Sources: Google Search Central, OpenAI, Anthropic, Perplexity. Checked 14 Nov 2026.

Google's statement is the clearest: for AI Overviews and AI Mode, you do not need AI text files at all.

Should you add an llms.txt file anyway?

It does little harm and takes an hour, so it is fine to add one if you like. Namova has one at namova.ai/llms.txt because it is a handy plain summary of the site for anyone, person or program, who wants it. Just do not expect it to change AI answers, and do not pay anyone to "optimise" it.

Put your time first into the things below, which the AI companies do document. The non-technical side, like reviews and being covered on other sites, is in how to get your business recommended by ChatGPT.

What is schema markup?

Schema markup, also called structured data, is a block of code (usually in a format called JSON-LD) that labels the facts on a page: this is a business, here is its address, these are its opening hours, this product costs this much. Google defines structured data as "a standardized format for providing information about a page and classifying the page content", and says "Google recommends using JSON-LD for structured data if your site's setup allows it" (Google Search Central).

The vocabulary comes from schema.org, a shared vocabulary for structured data. Common types for businesses include LocalBusiness, Organization, Product, Service and FAQPage.

Does schema markup help ChatGPT and other AI apps?

It helps machines read your facts without guessing, which is useful. But it is not a switch for AI answers, and the AI companies do not say they reward it:

  • Google says "there's also no special schema.org structured data that you need to add" for its AI features, and that structured data should match the visible text on the page (Google Search Central).
  • Google also says it "does not guarantee that your structured data will show up in search results, even if your page is marked up correctly" (Google structured data policies).
  • OpenAI, Anthropic and Perplexity document their search bots but publish nothing about how they use schema markup.

So add accurate schema for your business, products and FAQs, because it is cheap and Google uses it for rich results. Do it after the page itself is clear, and never mark up facts the page does not show.

What technical checks matter more?

These are the checks the AI companies actually document, in order of importance.

  1. Allow the AI search bots in robots.txt. OpenAI: sites that block OAI-SearchBot "will not be shown in ChatGPT search answers" (OpenAI). Anthropic: blocking Claude-SearchBot "may reduce your site's visibility and accuracy in user search results" (Anthropic). Perplexity recommends allowing PerplexityBot to appear in its results (Perplexity).
  2. Decide training separately from search. GPTBot and ClaudeBot are the training bots. You can block them and still allow the search bots. Google's equivalent control is Google-Extended, which covers Gemini training and grounding (giving Gemini content from Google Search to answer from); Google says it "does not impact a site's inclusion in Google Search" (Google crawler documentation). AI Overviews and AI Mode use pages that are indexed and eligible for normal Search (Google Search Central), so it follows that blocking Google-Extended should not remove you from them.
  3. Check your firewall and hosting too. Google reminds owners that crawling must be allowed in robots.txt and "by any CDN or hosting infrastructure" (Google Search Central).
  4. Make sure Google can index your important pages, since Google's AI features need pages that are indexed and eligible to appear in Search.
  5. Put your key facts in the page text, not only in images, PDFs or scripts.
  6. Add accurate schema for your business, products and FAQs.
  7. Then, if you want, add llms.txt.

A simple robots.txt that allows the AI search bots looks like this, written as plain lines in the file:

  • User-agent: OAI-SearchBot, then Allow: /
  • User-agent: Claude-SearchBot, then Allow: /
  • User-agent: PerplexityBot, then Allow: /

Ask whoever looks after your website to add or check these lines, and to make sure no other rule blocks them.

Where does Namova fit?

Namova checks how often AI apps name you for your customers' questions, then writes the changes worth making, including technical fixes like these when they are the problem, and checks again. Start with a website snapshot, which includes a readiness check of your website, or read the GEO vs SEO guide for the bigger picture.

Frequently asked questions

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

robots.txt is an established file that tells automated visitors which pages they may read, and the AI companies document how their bots follow it. llms.txt is a proposed summary file for AI tools that the major AI companies have not said they use for answers.

Is llms.txt a standard?

No. Its own site calls it a proposal, open for community input.

Which schema types should a local business use?

LocalBusiness (or a more specific type such as Plumber or Dentist), with your name, address, phone, hours and area served, plus FAQPage if you have a visible FAQ. Keep every value identical to what the page shows.

Can schema markup hurt my site?

Markup that does not match the page, or describes things that are not there, breaks Google's structured data guidelines and can make your pages ineligible for rich results. Accurate markup is low risk.

Do I need a developer for this?

For robots.txt and schema, usually whoever looks after your website can do it in an hour or two. Many website builders have settings or plugins for both.

Sources

Every link was checked on 14 November 2026. Namova is independent and not affiliated with any company named here.

See where you stand

A website snapshot asks Perplexity, OpenAI and Claude six questions your customers might ask, once each, and shows who gets named. One ask per question is a first look, not a measurement.

Run a website snapshot

No guarantee of results. Namova never promises rankings or what an AI app will say.