Does llms.txt Help You Show Up in ChatGPT? What Our Server Logs Say

It costs little to add, but it is not the lever. In 15 days our llms.txt received 93 requests, none of them from ChatGPT, Claude, Perplexity or Google. Over the same days, ChatGPT's search crawler read 328 regular pages on kiwop.com. It reads your pages, not your llms.txt.

See our GEO service See the 300-store study
  • 0 of 93 Requests to our llms.txt that came from ChatGPT
Scroll

Definition

Does llms.txt help you show up in ChatGPT?

Not measurably. llms.txt is a text file that summarises a website for language models, proposed by Jeremy Howard in September 2024. In the kiwop.com logs from 11 to 25 September 2026 it received 93 requests and none from GPTBot, OAI-SearchBot or ChatGPT-User, while OAI-SearchBot read 328 pages of the site. Google states in its official guide that Search does not use it. Having one does no harm; it is not what gets you into ChatGPT.

Where it does get used is technical documentation: the spec itself explains that coding agents follow it to find API references. For an online store or a services website, what decides whether ChatGPT cites you is whether its search crawler can read your pages and whether those pages answer specific questions with data. We have had an llms.txt since February 2026 and keep it up to date because it takes minutes, but we do not invest any more in it.

In numbers

What We Saw in Our Logs and in 300 Online Stores

Our own data, September 2026. The source of every figure is listed at the end of the page.

  • 1 of 93 Requests from a declared bot To kiwop.com/llms.txt over 15 days. The only one was Common Crawl's CCBot, on 12 September.
  • 0 From ChatGPT, Claude, Perplexity or Google Not GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot, Googlebot or Bingbot.
  • 328 Pages read by OAI-SearchBot ChatGPT's search crawler, IP-verified, over the same 15 days.
  • 13% Stores with their own llms.txt Out of 300 Spanish online stores. Another 13% have the template Shopify adds on its own.

Key points

What Actually Moves the Needle in ChatGPT

Three things, ranked by impact; llms.txt comes after them. The step-by-step guide is in how to get cited by ChatGPT.

  1. 01

    Let the crawler in

    If your robots.txt blocks OAI-SearchBot, you do not appear in ChatGPT search answers.

  2. 02

    One page per question

    With the answer at the top and data of your own, like the data on this page.

  3. 03

    Get others talking about you

    A mention in an industry publication or comparison carries more weight than anything you publish on your own domain.

Who it is for

Should You Have an llms.txt?

Adding one costs little. The real question is what you expect from it.

Who it's for

  • You have technical docs, an API or an SDK: coding agents do follow it to find references, and that is the use its own spec highlights.
  • You already have one and just want to maintain it: a quick review whenever your services or prices change, so it never says anything false.
  • Your store runs on Shopify: you already have one you did not write. Read it and decide whether to replace it with your own.
  • Your site is used by agents browsing on someone's behalf (bookings, purchases, support): Chrome includes it in Lighthouse's Agentic Browsing category. If you serve one, make sure it never returns a server error, because that does fail the audit.

Who it's not for

  • You run an online store or a services site and expect llms.txt to get you into ChatGPT: no data supports that.
  • You want to rank in Google: its official guide says Search ignores it.
  • Someone wants to sell you an llms.txt generator or consultancy: the file takes under an hour, and the money works harder in content and sources.
  • You block OAI-SearchBot in robots.txt and think llms.txt makes up for it: with the crawler locked out, there is nothing to make up for.

FAQ

llms.txt: Frequently Asked Questions

What people ask us after someone told them to add one.

Does ChatGPT read llms.txt?

Not in our data. From 11 to 25 September 2026, neither GPTBot, OAI-SearchBot nor ChatGPT-User requested kiwop.com's llms.txt, while OAI-SearchBot read 328 pages of the site. OpenAI's crawler documentation explains how to appear in ChatGPT through robots.txt and says nothing about other sites' llms.txt. And Ahrefs' study of 137,210 domains found that 97% of llms.txt files received not a single request in May 2026.

Does Google use llms.txt?

No. Its official guide to optimising for AI in Google Search (updated 10 July 2026) says you do not need AI text files to appear in Search, because Google Search does not use them, and that having one neither helps nor hurts. Back in June 2025, Google's John Mueller wrote that no AI system was using it.

What about Claude or Perplexity?

In our logs, neither ClaudeBot nor PerplexityBot requested it in 15 days. Anthropic's help centre explains how to control its three crawlers with robots.txt alone, with no mention of llms.txt. We know of no official statement from Perplexity saying it uses the file.

So should I remove my llms.txt?

No need. Keeping it costs almost nothing and Google confirms it does no harm. kiwop.com has had an llms.txt since February 2026 and we update it when something important changes, like a price or a service. What we do not do is spend more time on it: in 15 days, only one declared bot asked for it.

My Shopify store has an llms.txt I never wrote. What is it?

It is the "Agent Instructions" template Shopify serves on its stores without anyone asking. In our study of 300 Spanish online stores, 38 of the 44 we identified as Shopify had it. Open yourdomain.com/llms.txt to see what it says about you, or run your store through our AI readiness checker.

What should an llms.txt contain if I decide to write one?

According to the original proposal: a heading with the site name, a short summary and lists of links to your key pages, each with a one-line description. Our advice: short and true, with what an assistant needs to get you right (what you do, where and, if they are public, your prices). A long template file adds nothing.

If llms.txt is not it, how do I show up in ChatGPT?

Make sure OAI-SearchBot can read your pages (open robots.txt, clean sitemap, zero errors) and that those pages answer specific questions with your own data. We explain it step by step in how to show up in ChatGPT.

Why do so many people recommend it?

Because it is easy to make and easy to sell, and publishing a file gets mistaken for someone reading it. It is certainly being published: Ahrefs found it on 28% of 137,210 domains, and we found it on 26% of 300 stores. But half of the stores' files were added by Shopify on its own, and in the Ahrefs study 97% went unrequested for a whole month.

Sources

Sources for This Page

Where every figure comes from. Captured on 25 September 2026.

  1. kiwop.com access logs (own data) 11 to 25 September 2026: 93 requests to /llms.txt, one from a declared bot (CCBot, 12 Sep) and none from GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot, Googlebot or Bingbot. OAI-SearchBot, with its IP checked against the ranges OpenAI publishes, read 328 pages in that period.
  2. Kiwop Labs: AI in Spanish ecommerce 2026 n = 300 stores, measured 1-2 September 2026, verified 5 September. 86 llms.txt files detected; 78 verified (26%): 39 written by the store (13%) and 39 Shopify templates in total (13%): 38 on stores identified as Shopify and 1 on a store with no platform detected. In other words, 38 of the 44 stores identified as Shopify had the template.
  3. Open study dataset (JSON, CC BY 4.0) Fields llms_txt_verificado_pct, llms_txt_propio_pct and llms_txt_plataforma_pct, plus the origin of each file per store.
  4. llmstxt.org: the original llms.txt proposal Jeremy Howard, 3 September 2024. File format, and its main use in software documentation for coding agents.
  5. Google Search Central: AI optimization guide Updated 10 July 2026. Google Search does not use AI text files, and having an llms.txt neither helps nor hurts.
  6. Search Engine Roundtable: Google says no AI system uses llms.txt 17 June 2025. Statement by John Mueller (Google) on Bluesky.
  7. Ahrefs: llms.txt study across 137,210 domains 15 June 2026, May 2026 data. 28% of domains publish an llms.txt and 97% of those files received no requests during the month.
  8. OpenAI: Overview of OpenAI Crawlers Sites that block OAI-SearchBot in robots.txt do not appear in ChatGPT search answers. It does not mention other sites' llms.txt as a signal.
  9. Anthropic: how to control its crawlers Three crawlers (ClaudeBot, Claude-User, Claude-SearchBot) controlled through robots.txt. No mention of llms.txt.
  10. Chrome for Developers: Lighthouse llms.txt audit Updated 5 May 2026. Part of the Agentic Browsing category: it fails if llms.txt returns a server error and does not apply if the file does not exist (404).

Next step

Showing Up in ChatGPT Takes Groundwork, Not a File

We check whether ChatGPT, Claude and Perplexity crawlers read your site, which questions in your sector they answer by citing others, and what content you are missing. Our GEO service starts at €1,500/month; if two technical fixes are all you need, we will tell you.

Request an AI visibility review
  • No commitment
  • Response in 24h
  • Custom proposal
Last updated: September 2026

Let's talk.

Initial technical consultation

AI, security and performance. Diagnosis with phased proposal.

  • NDA available
  • Response <24h
  • Phased proposal

Your first meeting is with a Solutions Architect, not a salesperson.

Request diagnosis