Guide · 8 min read

llms.txt: what it is, how to create one and who reads it

An llms.txt file is a short markdown document at the root of your site that tells language models what the site is and which pages matter most. It takes an afternoon to write. The harder question is who actually reads it, and the honest answer in 2026 is: fewer systems than many SEO blogs suggest. Here is the format, a worked example, the real state of adoption and where it fits in a GEO plan.

See plans and pricing

llms.txt: what it is, how to create one and who reads it

What llms.txt is

The proposal comes from Jeremy Howard, co-founder of Answer.AI, who published it in September 2024 on llmstxt.org. The idea is simple: web pages are full of navigation, scripts and layout that a language model has to wade through, and its context window is limited. A plain markdown file at /llms.txt gives a model a clean summary and a curated list of links instead.

It is a proposal, not a standard. No standards body has adopted it, and no search engine or AI assistant is obliged to look for it. That matters for everything that follows.

The same proposal also suggests offering markdown versions of important pages at the same address with .md added, so that an agent can fetch the text without the HTML around it.

The llms.txt format, section by section

The file lives at the root of the domain, for example example.com/llms.txt, and is written in markdown. The specification defines a fixed order:

  1. An H1 with the name of the site or project. This is the only required part.
  2. A blockquote with a short summary: what the site is, who it serves and the key facts a reader needs to understand the rest.
  3. Optional paragraphs or lists with more detail. No headings here.
  4. One or more H2 sections, each with a list of links in the form [name](url), optionally followed by a colon and a short note.
  5. An optional H2 called “Optional”, for secondary links that a model can skip when it needs a shorter context.

You will also see llms-full.txt mentioned. It is not part of the core proposal: it is a convention adopted by several documentation tools, where the full text of the important pages is concatenated into one large file. It makes sense for developer documentation; for a company website it is rarely needed.

How to create an llms.txt file in 5 steps

  1. Pick the pages that matter. Usually 10 to 30: home, main services or product categories, pricing, the key guides, contact and policies a customer would ask about. Leave out tag pages, thin posts and anything you would not want quoted.
  2. Write the summary. Two or three sentences with facts, not slogans: what you do, for whom, where, and what makes you different in verifiable terms (prices, coverage, certifications).
  3. Group the links under H2 sections that a person would understand: Services, Pricing, Guides, Company. Add a short note after each link when the title alone is ambiguous.
  4. Use full URLs, check that every link returns a 200 status code and that none is blocked in robots.txt or behind a login.
  5. Upload the file as plain text to the root of the domain, open yoursite.com/llms.txt in a browser to confirm it loads, and add a reminder to update it whenever pricing or key pages change.

A realistic llms.txt example

This is an invented example for a fictional dental clinic. In the real file each link uses the full address with https://; here they are shortened so the page stays readable:

  • # Northwind Dental Clinic
  • > Private dental clinic in Leeds, UK, open since 2009. General dentistry, implants and orthodontics for adults and children. Evening appointments Monday to Thursday.
  • ## Treatments
  • - [Dental implants](northwind-dental.example/implants.md): single and full-arch implants, consultation and CT scan included.
  • - [Invisalign](northwind-dental.example/invisalign.md): clear aligners for adults and teens.
  • ## Practical information
  • - [Prices](northwind-dental.example/prices.md): fees for every treatment and payment options.
  • - [Contact and location](northwind-dental.example/contact.md)
  • ## Optional
  • - [Team](northwind-dental.example/team.md)

Our own file follows the same pattern: you can read it at recommendedgeo.com/llms.txt. It has a summary, our three plans with their prices and a list of the main pages in the six languages we work in.

Who reads llms.txt today

This is where most articles overpromise. Here is what can be checked as of late 2026:

  • Google: its Search documentation on AI features says you don’t need to create new machine-readable files or AI text files to appear in AI Overviews or AI Mode. Gary Illyes said in 2025 that Google does not support llms.txt, and John Mueller has compared it to the old keywords meta tag.
  • Chrome: Lighthouse added an experimental check for llms.txt in 2026, inside a category about how well browser agents can use a site. It sits apart from the SEO audits and says nothing about rankings or citations.
  • OpenAI, Anthropic and Perplexity: none of them has published documentation saying their crawlers or answer engines use llms.txt to choose sources. Their documented crawler controls are in robots.txt.
  • Developer tools: the clearest real use is in documentation. Anthropic, for example, publishes an llms.txt for its developer docs, and coding assistants can read such a file when a developer points them to it.

Several SEO tools and studies report server logs where AI bots rarely or never request the file. We have not seen a public, reproducible dataset that settles the question either way, so we do not quote figures. The safe conclusion: an llms.txt file does not, by itself, get your brand mentioned in ChatGPT or in AI Overviews.

When llms.txt is worth doing

Given that, should you bother? In most cases the answer is yes, as long as it takes an hour or two and nobody sells it to you as a ranking factor.

  • Worth it: software and API companies with technical documentation, because developers do feed these files to coding assistants.
  • Worth it: sites where agents may need clean facts, such as pricing, opening hours or product specifications, and where the HTML is heavy.
  • Reasonable: any company that already has its basics in order, as a low-cost bet on a convention that may or may not spread.
  • Low priority: a site whose pages are blocked to AI crawlers, have no clear answers or lack outside mentions. Fix those first.

Common mistakes with llms.txt

  • Treating it as a robots.txt replacement. It does not allow or block anything; crawler access is controlled in robots.txt, which we cover in robots.txt for AI bots.
  • Listing every URL in the sitemap. The value of the file is the selection; hundreds of links defeat the purpose.
  • Writing marketing copy in the summary. A model gets more from “orthodontics clinic in Leeds since 2009” than from “the leading smile experience”.
  • Linking to pages that redirect, return errors or are noindexed.
  • Saying something in llms.txt that the site itself does not say, such as a price or a service. Inconsistent facts make a source less trustworthy.
  • Publishing it once and forgetting it while prices and pages change.

How llms.txt fits into a wider GEO effort

If you think of GEO as making your brand easy to find, understand and recommend for AI systems, an llms.txt file is a small item on the “easy to understand” side. The items that carry more weight are elsewhere:

  • Access: AI crawlers allowed in robots.txt and pages that render without JavaScript tricks.
  • Structured data: schema.org markup for your organization, services and products that matches the visible text.
  • Citable content: pages with direct answers, concrete data and one topic each, as explained in how to write content AI cites.
  • Outside presence: mentions in comparisons, directories, reviews and press, which is often what decides a recommendation; see brand mentions for AI.
  • Measurement: running your customers’ real questions in each assistant every month, as described in how to measure AI visibility.

We create and maintain the llms.txt file as part of the technical work in every plan, next to robots.txt and structured data, and we keep it consistent with the rest of the site. Plans start at $449 per month with a three-month minimum; the details are on pricing. If you would rather see first where your site stands, ask for a free GEO audit.

Frequently asked questions

Does llms.txt help me rank in Google?

No. Google’s Search documentation says no special AI text files are needed to appear in AI Overviews or AI Mode, and Google representatives have said Search does not use llms.txt. Normal SEO work is what counts there; see how to appear in AI Overviews.

Does ChatGPT read llms.txt?

OpenAI has not published any documentation saying its crawlers or ChatGPT search use llms.txt to select sources. A model may read the file if a user gives it the URL, the same as any other page.

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

llms.txt is a short index: a summary plus links. llms-full.txt is a convention used by documentation tools that puts the full text of the key pages in a single file. Most company websites only need the first one.

Can llms.txt block AI crawlers?

No. It contains no rules. To allow or block bots such as GPTBot, ClaudeBot or PerplexityBot you use robots.txt.

Where do I put the llms.txt file?

At the root of your domain, so it loads at yoursite.com/llms.txt, served as plain text. On WordPress you can upload it by FTP or generate it with an SEO plugin, and then check the result by hand.

Let us handle llms.txt and the rest of your technical GEO

See plans and pricing
The four workstreams of GEO: measuring AI answers, citable content, outside presence and technical GEO
AI assistants tracked every month: ChatGPT, Gemini, Perplexity, Claude, Copilot and Google AI Overviews
Monthly GEO process: measure, fix, publish and report

Pay with Stripe

On the live site this button takes you straight to secure card payment (Stripe). This prototype doesn't charge yet.

Scroll to Top