14.3% of the Top 1 Million Websites Now Serve an llms.txt File, but Only 7.9% of the Top 1,000 Do

Summary

Picture a bakery owner who reads that AI assistants look for a file called llms.txt. She checks her own site and finds nothing there. Then she wonders whether she is behind, or whether almost nobody has one.

llms.txt is a plain text or markdown file placed at the root of a website. It gives AI tools a short summary of the site and a list of the pages that matter most. No search engine or AI company requires it, but it has become the most talked-about AI SEO file of the past year.

We measured it across 772,412 of the world's most visited homepages. 110,188 of them (14.3%) now answer /llms.txt with a text file. Most of that comes from online stores whose platform adds the file for them, and the most popular sites are less likely to have one than the long tail.

What we measured

We used the HTTP Archive October 2026 mobile crawl in BigQuery (httparchive.crawl.pages) and kept homepages of sites ranked in the Chrome UX Report top 1 million. HTTP Archive requests /llms.txt for every site and records the HTTP status and content type in a custom metric called llms_txt_validation.

We counted a site as having an llms.txt file when /llms.txt returned a 200 response served as text/plain or text/markdown. HTML responses, 404s, blocked requests and timeouts do not count. Platform comes from HTTP Archive's technology detection, using one primary platform per site.

Adoption is higher in the long tail than at the very top

Rank buckets come from the Chrome UX Report. The top 1K bucket covers the 1,000 most visited origins, 738 of which are in this crawl; the last bucket holds sites ranked 500,001 to 1 million.

The top 10,000 sites sit at 9.1% (662 of 7,244). The 100K to 500K bucket reaches 13.4% and the 500K to 1M bucket 15.7%. The lower buckets hold far more small online stores, and that is where automatic files show up.

A handful of platforms publish the file for almost every site they host

This table names platforms only to show who uses the file. It says nothing about file quality.

Seven platforms sit above 95%: Shopify, Shoptet, Tiendanube, Shopline, Nuvemshop, Wix and Duda. Together they account for 50,438 of the 110,188 files (45.8%). Site builders like Webflow (24.3%) and frameworks like Astro (25.1%) land in the middle, and Squarespace sites rarely have one (4.0%).

Without automatic files, about 1 in 13 top sites has an llms.txt

We removed the seven platforms that add the file by default and recounted.

Once automatic files are set aside, the curve flattens. Every one of the 58 files in the top 1,000 sits outside those platforms, which suggests large sites that have one chose to add it.

What this means

If your site runs on a platform that already publishes llms.txt, check what it says. An automatic file lists whatever the platform thinks matters, which may not be your best service pages or your contact details. If you run WordPress or a custom site, adding one is a short job, and you are still ahead of roughly 86% of popular sites. Either way, the file is a small part of AI search optimization, so keep the time you spend on it short.

Keep expectations realistic. The file is a hint for AI tools, not a ranking switch, and the biggest sites on the web mostly do without it. Treat it as one tidy signal next to the basics: crawlable pages, clear facts about your business and content that answers real questions. Want to know if AI assistants can find and quote your business right now? Our free AI visibility audit tests it with real prompts.

Limits

HTTP Archive checks only the homepage origin's /llms.txt; files on other paths or subdomains are not counted.

We count a 200 response with a text/plain or text/markdown content type. We did not read file bodies, so a small number of empty or placeholder files may be included.

Platform detection relies on HTTP Archive technology fingerprints; some sites go undetected and fall into 'Other or none detected'.

Rank buckets are approximate popularity groups from the Chrome UX Report, not exact traffic ranks.

Adoption is not proof that any AI system reads or uses the file.

Frequently asked questions

Related studies

• llms.txt and AI instruction filesllms.txt Now Outnumbers security.txt Nearly 5 to 1
• llms.txt and AI instruction filesllms.txt Grew From 1.1% to 14.3% of Top Sites in 15 Months
• llms.txt and AI instruction files46% of llms.txt Files Are Added Automatically by Platforms
• llms.txt and AI instruction files15.6% of Top Sites Say 'Found' for Pages That Don't Exist