llms.txt is a plain-text markdown file placed at a page’s own /llms.txt address, meant to give AI systems a compact, structured summary of that page without forcing them to parse a full, cluttered HTML document. It was proposed in September 2024 by Jeremy Howard, co-founder of the AI research lab Answer.AI, as a community standard rather than an official web protocol.
Search engines have had a settled file for decades: robots.txt tells a crawler what it can and cannot access. Nothing equivalent existed for telling an AI system what a page actually contains, in a format built for a language model to read rather than a browser to render. llms.txt is an attempt to fill that specific gap.
This guide covers what the file actually contains, why it exists, and what it means for a creator’s bio page specifically, including how much weight to put on it.
Key Takeaways
- llms.txt is a proposal, not an official standard. It has no formal standards body backing it and no committed adoption from major AI search engines.
- The format is deliberately minimal. One H1 title, a one-line blockquote summary, and H2 sections linking to more detailed resources.
- It solves a real, specific problem. Language models have limited context windows, and converting a full HTML page with navigation, ads and scripts into clean text is difficult and imprecise.
- It is not a ranking or citation guarantee. Publishing one does not make an AI system more likely to cite your page over another.
- It costs almost nothing to have. A page that generates it automatically loses nothing by shipping it, even while its actual impact on citation is still unproven.
Where llms.txt Came From
Jeremy Howard published the original /llms.txt proposal on llmstxt.org in September 2024, describing the problem plainly: large language models increasingly rely on website information to answer questions, but their context windows are too small to process most websites in full, and converting a complex HTML page, with its navigation, ads and JavaScript, into clean, LLM-friendly text is both difficult and imprecise (llmstxt.org, accessed September 2026).
The proposal has no formal standards body behind it. It is a community convention, published and maintained independently, with no confirmed adoption commitment from OpenAI, Google, Anthropic or Perplexity as an official input to their systems (llmstxt.org, accessed September 2026). That distinction matters: llms.txt is closer to an early, hopeful convention than a settled protocol every AI system is known to consume.
What’s Actually Inside an llms.txt File
The specification is intentionally narrow. A compliant file contains:
| Section | Purpose |
|---|---|
| H1 title | The name of the site or page |
| Blockquote summary | A short, one or two sentence description of what the page is |
| Body text (optional) | Additional context in plain prose, no headings |
| H2 sections | Grouped links to more detailed resources, each with a short description |
| Optional section | Lower-priority links that can be skipped if context is tight |
The related file, /llms-full.txt, goes further: a single, comprehensive plain-text version of a site’s full documentation or content in one place, for systems that can handle more context.
What llms.txt Is Not
It is not a replacement for structured data. Structured data (JSON-LD) describes specific facts on a page in a standardized, machine-parseable format, a product’s price, an event’s date. llms.txt is a plain-language summary, closer to an executive briefing than a database entry.
It is also not a search ranking factor, and it is not a guarantee of AI citation. Its purpose is narrower: giving an AI system a fast, accurate way to understand what a page is about, if that system chooses to fetch and use the file at all.
What a Minimal llms.txt Actually Looks Like
Stripped down to the format’s own rules, a compliant file for a creator page reads close to this:
# Jamie Rivera
> Photographer and preset creator based in Austin. Shares editing tutorials
> and sells Lightroom preset packs.
## Links
- [Preset shop](https://example.com/shop): Lightroom preset packs by category
- [Booking](https://example.com/book): Session booking calendar
- [About](https://example.com/about): Full bio and contact detailsOne title, one summary line, a short list of links with a description each. Nothing about the format calls for more than that, which is why it costs so little to generate automatically alongside a page’s other content.
Does llms.txt Actually Help Get You Cited?
Honestly, the evidence for llms.txt directly driving AI citations is thin, and any AI-readability guide that oversells it is doing you a disservice. It is one small piece of a much larger picture that includes structured data, clean semantic HTML, and a per-page policy on whether AI crawlers can access the page at all. More.You keeps that last piece with the page owner, one page at a time, through its AI crawler controls.
Where it earns its place is cost, not proven impact. Generating it automatically from a page’s own content costs nothing to maintain and does not compete with anything else for effort. Treat it as a low-cost, plausible-upside addition, not the centerpiece of an AI visibility strategy.
What This Means for Your Bio Page
An AI-native link in bio page ships llms.txt as one layer among several: alongside an agent manifest, structured data generated from the page’s actual blocks, and a stated AI-crawler policy the page owner controls. None of these layers does the whole job alone. Together, they give an AI system several independent ways to understand your page correctly.
If you are checking whether your own page currently ships anything like this, making your link in bio AI-readable covers the practical audit, and answer engine optimization covers the wider set of factors that actually influence whether an AI system cites you.
Your bio page’s llms.txt should be generated the moment you publish, from the same header, bio and blocks a visitor sees. Create your More.You page and every page ships its own llms.txt automatically, no separate file to maintain by hand.
FAQ
Who created llms.txt?
Jeremy Howard, co-founder of the AI research lab Answer.AI, published the original proposal at llmstxt.org in September 2024 (llmstxt.org, accessed September 2026).
Is llms.txt an official web standard?
No. It is a community-proposed convention with no formal standards body backing it, similar in spirit to robots.txt but without the decades of established, universal adoption that file has.
Do I need to write my own llms.txt file by hand?
Not if your website or bio page tool generates one automatically from your existing content. Manually maintaining a separate summary file alongside your real content is extra work most creators do not need to take on.
Does having llms.txt guarantee ChatGPT or Perplexity will cite my page?
No. It has no confirmed adoption commitment from major AI systems as an official input, and no evidence currently shows it reliably drives citation on its own. It is one small, low-cost piece of a broader AI-readability setup.
What is the difference between llms.txt and robots.txt?
robots.txt tells crawlers what they are and are not allowed to access on your site. llms.txt is a content summary meant to help an AI system understand a page faster once it has access. One is a permission file, the other is closer to a briefing document.
Sources
The /llms.txt file (llmstxt.org, accessed September 2026). More.You product brief, section 3 (the AI-native layer).
