Every few months, a new “must do” tactic for AI search goes around marketing circles. In 2025 and 2026, one of the loudest was llms.txt: a small text file that supposedly helps ChatGPT, Perplexity and Google’s AI features understand your site and cite it more often.
If you run a company blog, it is a tempting idea. One file, ten minutes of work, and maybe more AI visibility. Plenty of SEO plugins and site builders now offer to create it for you with one click.
But does it actually work? Google has now answered in writing, and several large studies have checked the server logs. Here is what the evidence says, and where your time is better spent.
Quick answer: No, llms.txt does not help your blog show up in Google Search, AI Overviews or AI Mode. Google’s official AI optimization guide says Google Search ignores llms.txt files, and an Ahrefs study of 137,210 domains found that 97% of llms.txt files received zero requests in May 2026. It does not hurt either. The file is mainly useful for developer documentation that AI coding tools read. For AI visibility, focus on crawl access, clear answer-first content and original information instead.
What Is llms.txt?
llms.txt is a plain Markdown file placed at the root of a website (for example, yoursite.com/llms.txt). It gives large language models a short, curated map of the site: what the site is, which pages matter most, and a one-line note about each one.
The format was proposed by Jeremy Howard, co-founder of Answer.AI, in September 2024. The idea is simple. Web pages are full of menus, scripts and cookie banners, so a clean summary could save an AI system time and tokens when it needs information from your site.
A basic file looks like this:
# Your Company Name
> One sentence about what your company does.
## Key pages
- [Pricing](https://example.com/pricing): Plans and prices
- [Product guide](https://example.com/guide): How the product works
- [Blog](https://example.com/blog): Articles and tutorialsTwo things llms.txt is not:
- It is not a robots.txt rule. It does not allow or block anything. If you want to control which AI crawlers can access your site, that still happens in robots.txt.
- It is not an official standard. It is a community proposal. No standards body backs it, and no major AI search product has committed to using it for ranking or citations.
What Google Says About llms.txt
Google has been unusually direct on this one. Its official guide, Optimizing your website for generative AI features on Google Search, has a “mythbusting” section that lists llms.txt as something you can ignore.
The guide says you don’t need to create “new machine readable files, AI text files, markup, or Markdown to appear in Google Search (including its generative AI capabilities), as Google Search itself doesn’t use them.”
It also says that creating an llms.txt file “will neither harm nor help your site’s visibility or rankings in Google Search, as Google Search ignores them.”
This position did not come out of nowhere. Here is the timeline:
- July 2025: Google’s Gary Illyes said Google doesn’t support llms.txt and had no plans to.
- May 2026: Google published its first AI optimization guide and listed llms.txt among the tactics you don’t need for AI Overviews and AI Mode.
- June 15, 2026: Google added clearer wording that machine-readable AI files are not used by Google Search and won’t help or hurt rankings.
- After that: Google’s John Mueller described llms.txt as “not done for search” and a “temporary crutch, perhaps to save some tokens” for AI coding tools reading developer docs, as reported by Ahrefs.
One source of confusion: Chrome’s Lighthouse tool added an experimental llms.txt check as part of its agentic browsing audits. That is a different team with a different goal (helping AI agents that browse sites). It is not a signal that Google Search uses the file.
What the Data Says: Do AI Crawlers Even Read It?
Google is only one AI search product. What about ChatGPT, Perplexity and Claude? This is where server log studies help, because they show whether AI bots actually request the file.
The biggest study so far comes from Ahrefs, published on June 15, 2026. It looked at 137,210 domains using Ahrefs Web Analytics and their bot traffic in May 2026. Key findings:
- 28% of the domains published an llms.txt file (Ahrefs notes this sample is more SEO aware than the average web, so treat it as an upper bound).
- 97% of those llms.txt files received zero requests in May 2026. No bots, no humans, nothing.
- Of the requests that did happen, only 19.5% came from named AI tools. GPTBot was first and Claude Code was second, ahead of every AI search and assistant bot.
- 12% of requests came from GEO tools, llms.txt checkers and researchers studying the file, not using it.
- Zero AI bots went looking for llms.txt on sites that didn’t have one.
Other studies point the same way. As summarized by Optimizely in October 2026:
| Study | Sample | What it found |
|---|---|---|
| Ahrefs | 137,210 domains | 97% of llms.txt files got zero requests in May 2026 |
| SE Ranking | About 300,000 domains | 10.13% had the file; no clear link to AI citations |
| Limy | 500M+ AI bot visits over 90 days | Only 408 requests went to llms.txt |
| Otterly | 62,100+ AI bot visits over 90 days | 84 requests touched llms.txt (about 0.1%) |
Different teams, different methods, same result. The bots that decide which pages get cited in AI answers mostly skip llms.txt and read your normal HTML pages.
Why AI Search Mostly Ignores llms.txt
Once you look at how AI search actually works, the results make sense.
- AI search pulls individual pages, not site summaries. When someone asks ChatGPT or Google AI Mode a question, the system retrieves the specific pages that answer it best. Google’s guide explains that its AI features rely on its core Search ranking systems and the Search index to find those pages.
- Crawlers already read HTML well. Removing menus and boilerplate from a page is a solved problem for modern crawlers.
- Self-written summaries are easy to game. A file where you describe your own site is a lot like the old keywords meta tag, which search engines stopped trusting for the same reason. John Mueller has made that comparison himself.
- No major AI search product has signed up. Without a big platform committing to read it, the file has no guaranteed audience.
In short, an llms.txt file tells AI systems what your site claims to be about. AI search would rather read the pages and decide for itself.
When llms.txt Is Worth Adding
llms.txt is not useless. It just has a smaller and more technical audience than the hype suggests. It can make sense if your company has:
- Developer documentation or an API. AI coding tools like Claude Code do fetch llms.txt to find the right docs page fast. OpenAI, for example, publishes an llms.txt file for its own API docs.
- A large help center or knowledge base. A curated index can point AI agents to the current page instead of an outdated one.
- Customers who use AI agents with your product. If agents act on your site or read your docs while helping users, the file has a real chance of being read.
If you do add one, keep the effort small. Many tools now generate it automatically. On WordPress, Rank Math includes an llms.txt option, and Ahrefs notes that Wix already creates these files for its users. Just make sure it updates when your content changes. A stale file with old prices or moved pages is worse than no file, because the few tools that read it may trust it.
Also, a small security point from the Ahrefs study: agents are built to trust this file, and bad actors are already probing llms.txt files for prompt injection. Treat it like any other public file on your site and keep an eye on what it says.
What to Do Instead to Show Up in AI Search
If your real goal is getting your blog posts mentioned and cited by ChatGPT, Perplexity, Gemini and Google’s AI Overviews, these steps matter far more than llms.txt.
1. Make sure AI crawlers can reach your pages
Check your robots.txt file and any firewall or CDN bot settings. If GPTBot, PerplexityBot, ClaudeBot or Googlebot are blocked, nothing else on this list matters. Our guide to optimizing the WordPress robots.txt file walks through the basics.
2. Put the answer first
Start each post, and each important section, with a direct answer in one or two clear sentences. Then add the details. This is easier for readers to skim and easier for AI tools to quote. For example, a post titled “How much does a CRM cost?” should state a real price range in its first paragraph, not after 500 words of background.
3. Publish information nobody else has
Google’s AI guide says unique, non-commodity content will likely influence your presence in AI search more than anything else. For a company blog, that means your own data, customer examples, real pricing, product comparisons from your team’s experience, and clear opinions. Generic “10 tips” posts that repeat what everyone says are easy for AI to replace.
4. Keep your structure clean
Use descriptive headings, short paragraphs, lists and tables where they help. You don’t need to chop content into tiny chunks (Google’s guide says that isn’t required either), but a page that is easy for a person to scan is also easy for an AI system to understand.
5. Keep facts current and dated
AI tools prefer sources that look up to date, especially for prices, tools and statistics. Update important posts when facts change and show a clear last updated date. We covered this in detail in Do Dates on Your Blog Affect SEO and AI Visibility?
6. Measure citations, not file requests
The only way to know whether your work is paying off is to track whether AI answers actually mention and cite your site. You can check by hand by asking ChatGPT or Perplexity the questions your posts answer. For ongoing tracking, OptimizeCamp (our own tool) runs your chosen prompts across ChatGPT, Google AI Overviews, Perplexity and Claude and shows where you are mentioned and cited. It is a newer product than some bigger names, so compare it with the other options in our list of the best AI visibility tools before you decide.
Frequently Asked Questions
Does llms.txt help with Google rankings?
No. Google’s official AI optimization guide says Google Search ignores llms.txt files, so having one will neither help nor harm your rankings, including in AI Overviews and AI Mode.
Does ChatGPT use llms.txt?
There is no public commitment from OpenAI that ChatGPT search uses llms.txt for citations. In the Ahrefs study, GPTBot was the most common named AI bot requesting llms.txt, but 97% of files still received no requests at all, and a fetch does not mean the file was used.
Can llms.txt hurt my SEO?
No. Google says the file has no effect on visibility in Google Search. The main risks are a stale file with outdated information and a file that someone tampers with, so keep it accurate if you publish one.
Is llms.txt the same as robots.txt?
No. robots.txt tells crawlers which parts of your site they may access. llms.txt is an optional summary file for AI tools and does not allow or block anything.
Should I delete my llms.txt file?
You don’t need to. If your SEO plugin or site builder already created it, leaving it in place costs nothing. Just make sure it stays accurate, and don’t spend more time on it expecting AI visibility gains.
What actually helps a blog get cited in AI search?
Crawl access for AI bots, answer-first writing, original information, clear structure and up-to-date facts. Then track AI mentions and citations so you can see which posts are working.
Final Thoughts
llms.txt is a reasonable idea that the AI search products most businesses care about have chosen to ignore. Keep it if your platform creates it automatically or if you run developer docs, but don’t expect it to get your blog cited.
Your time is better spent on crawlable, answer-first, original content that AI tools want to quote. For a practical checklist, read our guide on how to get your blog posts cited in AI search.