If you are going to invest time in implementing a new technical standard, you probably want to know whether anyone else is using it and whether it is actually working. For llms.txt, the honest answer is nuanced. A growing list of companies have published llms.txt files, but there is no published research, no case study, and no public data showing that the file has improved AI visibility, traffic, or citations for any of them.
In this guide, we examine real adoption patterns, look at the companies that have implemented llms.txt, analyze what their files actually contain, and explain what this all means for your site. We also cover why the lack of proven results does not necessarily mean the standard is worthless, and why early adoption might still be a smart move for the right kind of website.
The goal here is not to hype llms.txt or dismiss it. The goal is to give you grounded, evidence-based information that helps you decide whether early adoption makes sense for your situation.
Pro Hint
When evaluating any emerging technical standard, look at who is adopting it first. The pattern of early adopters tells you which use cases the standard was designed for and which ones it genuinely serves. For llms.txt, the early adopters are almost all documentation platforms and developer tool companies with highly structured content. If your site is in that category, you are the target user.
Who Has Actually Implemented llms.txt
The most comprehensive public directory of llms.txt adopters is maintained at llmstxt.cloud. As of mid-2026, the directory lists roughly two dozen organizations that have published llms.txt files publicly accessible at the root of their domains. The list is not massive, but it is growing, and the names on it are informative.
Documentation Platform Early Adopters
Mintlify, a popular documentation hosting platform, publishes an llms.txt file at its root domain. Their file is well-structured, with sections for product documentation, API references, changelog, and getting started guides. Each section includes concise descriptions of the resources. This is exactly what an effective llms.txt file should look like: organized, descriptive, and useful to an AI system encountering it for the first time.
The reason documentation platforms lead adoption is straightforward. Their entire business model depends on content being referenced and cited correctly. An AI assistant that summarizes their documentation accurately drives qualified traffic and builds trust in their product. If an llms.txt file helps with that, it is a low-cost addition with potential high upside.
Developer Tools and Infrastructure Companies
Tinybird and Cloudflare have both published llms.txt files. Both companies produce technical content that developers reference daily when building applications. Tinybird’s file guides AI systems to their real-time data API documentation, tutorials, and blog posts. Cloudflare’s file points to their developer docs, changelog, and security resources. In both cases, the content is highly structured, frequently updated, and valuable for AI training and retrieval.
The pattern is clear. Companies whose content is referenced by developers using AI coding assistants benefit most from llms.txt. When a developer asks an AI assistant how to use a specific API or integrate with a service, the assistant needs to find the right documentation page quickly and accurately. llms.txt provides a navigation aid that speeds up that process.
The Anthropic Connection
Anthropic, the company behind Claude, publishes its own llms.txt file at anthropic.com. This has generated significant attention because it might imply that Claude respects llms.txt during crawling or training. The reality, as of mid-2026, is more modest. Anthropic has not publicly stated that their crawlers or models use llms.txt as a retrieval or indexing signal. They publish it as a community gesture, to demonstrate best practices and encourage ecosystem adoption.
Whether Anthropic’s crawlers consult the file internally is not publicly documented. What is clear is that Anthropic’s adoption gives llms.txt credibility that no other organization can match. When the company that proposed the standard on llmstxt.org also publishes a compliant file, it signals serious intent to the broader ecosystem.
Why These Companies and Not Others
Look at the companies that have adopted llms.txt and a pattern emerges. They are all technology companies with deep, structured, publicly referenced content. They have developer audiences who use AI tools daily. Their content is stable enough to curate but dynamic enough to benefit from ongoing curation. They have technical teams capable of creating and maintaining the file.
Contrast this with companies that have not adopted llms.txt. Major e-commerce platforms, news organizations, and entertainment companies are largely absent. This is not because they do not care about AI visibility. It is because their content structures are less conducive to the llms.txt format. A product catalog page, a news article, and a streaming show page do not benefit as much from the kind of structural curation that llms.txt provides.
What the Adoption Data Actually Shows
Two dozen adopters sounds like progress until you consider that there are over 200 million active websites on the internet. Even among the subset that would benefit from llms.txt, adoption is in the single-digit percentages. The raw numbers tell a clear story: llms.txt is an early-stage protocol with enthusiastic but limited adoption.
What the adoption data also shows is that the companies who have implemented llms.txt tend to be thoughtful about it. The files on the llmstxt.cloud directory are generally well-structured, accurately maintained, and professionally written. They are not placeholder files stuck up as an afterthought. This suggests that the early adopters see genuine value in the format, even if they cannot yet prove measurable results.
The Case Study Gap
Here is what is missing from the adoption narrative: data. No company on the llmstxt.cloud directory has published a case study showing that llms.txt improved their AI citations, search visibility, or organic traffic. No SEO tool has measured a correlation between llms.txt presence and ranking improvements. No AI researcher has documented that llms.txt influences model behavior during training or inference.
This gap matters because it means the value of llms.txt is currently theoretical. The logic is sound: if AI systems start using the file, sites with good files will benefit. But logic is not evidence. Until a major AI provider announces support for llms.txt and we see measurable results from early adopters, the standard remains unproven.
Real-World Examples of Effective llms.txt Files
Even without measurable results, we can learn from the files that have been published. Examining the structural choices of effective llms.txt files teaches us how to build better ones for our own sites.
What Makes an Effective llms.txt File
Effective llms.txt files share several characteristics. They use descriptive section headers that an AI system can use to build a mental model of the site. They include one-line descriptions for most links, providing context about what each resource contains. They are organized hierarchically, with the most important resources listed first within each section. They avoid generic labels and vague descriptions. And they are kept current, with broken links removed and new resources added as the site evolves.
Ineffective llms.txt files share a different set of characteristics. They are long flat lists of links with no descriptions. They include every URL on the site rather than a curated selection. They use generic section names like “Links” or “Resources.” They contain broken links and outdated URLs. And they were created once and never updated.
Analyzing a Real llms.txt File
Let us say you visit the Cloudflare llms.txt file. You see sections for “Getting Started,” “Products,” “Resources,” and “Network.” Each section has descriptive links with context. The “Getting Started” section tells you where to find setup guides and quickstart tutorials. The “Products” section lists product categories with one-line descriptions. The file is scannable, organized, and genuinely useful for an AI system trying to understand Cloudflare’s content architecture.
Now compare this to a hypothetical bad llms.txt file: a flat list of 300 links organized into three vaguely labeled sections with no descriptions. An AI system reading this file would get almost no useful signal. It would know there are three sections and roughly how many links each contains, but it would not know what any of the links actually contain or why they matter. This is the difference between a thoughtful implementation and a checkbox exercise.
The Criticism: llms.txt as a Solution Looking for a Problem
It would be irresponsible to present only the optimistic view. A legitimate and widely held criticism of llms.txt is that it solves a problem that does not yet exist for most website owners. The argument goes like this: AI crawlers already find and cite good content through traditional crawling. The established tools, robots.txt, XML sitemaps, and schema markup, already provide sufficient guidance. Adding llms.txt is unnecessary complexity for a benefit that has not been demonstrated.
This criticism has merit. If your primary goal is to improve your SEO or AI citations in the next six months, llms.txt will not help you. There are no proven results, no official support, and no guarantee that the standard will gain the adoption needed for it to matter. Investing significant resources in llms.txt right now is not pragmatic for most sites.
Why the Criticism Does Not Tell the Whole Story
The counterargument is that standards development moves faster than most people expect. XML sitemaps went from obscure proposal to essential SEO tool in under a decade. Open Graph tags became mandatory for social sharing in a similar timeframe. The timeline is compressed now, with AI development moving at speeds that make three to five year adoption windows feel optimistic rather than conservative.
For sites with the right profile, the cost of being an early adopter is near zero, especially with automated tools like Yoast SEO’s llms.txt generator. The potential upside, if the standard gains traction, is significant. When adoption is cheap and the potential payoff is large, early adoption becomes a rational choice even without guaranteed results.
What the Real-World Results Tell Us About the Opportunity
Here is the practical conclusion from the adoption data. llms.txt is real, it is growing, and it has attracted thoughtful early adopters. But it is not a proven driver of AI visibility, traffic, or citations. The opportunity is forward-looking, not immediate. The value depends on future AI platform adoption that has not happened yet.
That makes llms.txt a bet, not a tactic. Bets are appropriate when the cost is low and the potential return is high. For documentation platforms, developer tools, and structured content sites, the cost of creating a good llms.txt file is measured in minutes. The potential return, if and when AI platforms adopt the standard, could be meaningful competitive positioning.
For other site types, the cost and benefit assessment works differently. If your content structure is simple and your audience does not use AI tools to find your content, llms.txt does not move any needles. Your time is better spent on the tactics that do produce results right now: creating content that AI systems can easily parse, implementing schema markup, and building authoritative backlinks.
For a complete analysis of which AI visibility tactics produce measurable results today and which are still waiting for their moment, check out our guide on what actually works for AI search visibility.
Frequently Asked Questions
As of mid-2026, roughly two dozen organizations are tracked on the llmstxt.cloud directory. This is a tiny fraction of the total number of websites on the internet, reflecting the early stage of llms.txt adoption. The number is growing but remains small relative to the potential addressable market.
No, there are no published case studies or public data showing that llms.txt has improved AI citations, search visibility, or organic traffic for any organization. The benefit remains theoretical until major AI providers officially adopt the standard.
No. Cloudflare publishes an llms.txt file for cloudflare.com, their own corporate website. Website owners who use Cloudflare as a CDN or hosting provider still need to create their own llms.txt file at their own domain if they want to participate in the standard.
Anthropic has not publicly confirmed whether their crawler or models use llms.txt files for indexing or training. Their adoption of the format appears to be a community gesture rather than a functional implementation. Do not assume that having an llms.txt file will make Claude cite your content more often.
Waiting guarantees you will not benefit from early mover advantages. If llms.txt becomes an official standard, the sites that already have well-structured files will be ahead of the curve. The cost of preparing now is low for most technical or content-heavy sites.
No peer-reviewed studies or published case studies exist on llms.txt effectiveness. This is a normal stage for an emerging technical standard, but it means that any claims about benefits are speculative rather than evidence-based. Evaluate llms.txt as a forward-looking bet rather than a proven tactic.