What Is Llms.Txt? Format, Examples, and Whether It Works

llms.txt is a proposed community standard for a plain-text markdown file at your site root that summarizes your website for large language models. Proposed by Jeremy Howard and Answer.AI in 2024, it is not an official directive and no major AI crawler has committed to honoring it. It is a low-cost, optional courtesy file, not a substitute for crawlable content.

TL;DR

• llms.txt is a proposed markdown file at your site root that gives LLMs a curated summary of what your site is about.
• It was proposed by Jeremy Howard and Answer.AI in 2024 as a community convention, not an official standard.
• No major crawler -- Google, OpenAI, or Anthropic -- has committed to honoring it today.
• It is not a replacement for crawlable content, structured data, or clean site architecture.
• Adding one is cheap, harmless, and may have upside as voluntary adoption grows.

What Is an Llms.Txt File?

An llms.txt file is a small markdown document that lives at the root of your domain, served at the path /llms.txt. The idea is that instead of an AI model having to crawl and parse your entire site to understand it, you provide a single, humans-readable summary that points the model to your most important pages. The format is intentionally simple: it is just a markdown file, nothing more.

The proposal specifies that the file begins with an H1 containing your site or project name. Immediately after, you may include a blockquote that gives a short summary of what the site is and who it is for. After the blockquote, you can add any amount of free-form markdown detail that explains context, conventions, or anything else a model might need. Then you organize the rest of the file into H2 sections, each containing a markdown list of links to key URLs with a short description of each.

The goal is curation. A model that reads your llms.txt should come away knowing your site name, a one-paragraph sense of what you do, and a short list of the pages that matter most. That is the entire ambition of the format. It is a courtesy, not a command.

How Is Llms.Txt Different from Robots.Txt and Sitemap.Xml?

These three files are easy to confuse because they all sit at the site root and all relate to how machines read your website, but they do very different jobs. Understanding the distinction is important, because conflating them leads to the false belief that llms.txt controls AI crawlers the way robots.txt controls search engines.

robots.txt is a directive file. It tells compliant crawlers which paths they are allowed or disallowed from fetching. Search engines largely honor it. sitemap.xml is a discovery file. It lists the URLs you want search engines to know about and helps them find content efficiently. Schema markup, by contrast, is structured data embedded in your pages that helps search engines and some AI systems understand the meaning of your content.

llms.txt sits in a different category entirely. It is neither a directive nor a discovery file in the technical sense. It is a proposal that a model, if it chooses to, can read to get a curated summary. It carries no enforcement mechanism. The table below makes the contrast explicit.

The key takeaway: robots.txt and sitemap.xml are infrastructure that crawlers rely on, while llms.txt is a suggestion a model may or may not read. If you want to control access, use robots.txt. If you want to be discovered, use sitemap.xml and schema. llms.txt is the optional layer on top.

What Does a Good Llms.Txt File Look Like?

A good llms.txt is short, honest, and genuinely useful as a table of contents for your site. It names the site, summarizes it in a sentence or two, and then links to the handful of pages a reader (human or model) most needs. It does not try to be a full copy of your site, and it does not stuff keywords.

Here is a realistic example for a fictional marketing consultancy. Notice the H1, the blockquote summary, and the H2 sections with link lists.

# Stackmatix

> Stackmatix is a growth marketing consultancy that helps B2B teams
> turn content and analytics into predictable pipeline. This file
> points LLMs to our most useful public resources.

## Primary Pages
- [Home](https://www.stackmatix.com/): Overview of services and approach.
- [LLM Optimization Guide](https://www.stackmatix.com/blog/llm-optimization-guide): How to make content legible to AI systems.

## Articles
- [AI Citations](https://www.stackmatix.com/blog/ai-citation-building-get-referenced-by-llms): Tactics for getting referenced by large language models.
- [LLM Search Optimization](https://www.stackmatix.com/blog/llm-search-optimization): Structuring content for AI-driven search.

## Resources
- [Pricing](https://www.stackmatix.com/pricing): Engagement models and rates.

That is it. The file is plain ASCII markdown, easy to read, and easy to maintain. The links should point to real, crawlable pages -- the value of llms.txt depends entirely on the pages it references being genuinely useful.

How Do You Create an Llms.Txt File Step by Step?

Creating the file is a straightforward task that takes most teams under an hour the first time. Here is the workflow.

1. Decide the single sentence that describes your site, and write it as the blockquote summary under the H1 site name.
2. List your five to fifteen most important URLs across your site: core product or service pages, top blog posts, and key resources.
3. Group those URLs into logical H2 sections such as Primary Pages, Articles, Documentation, or Resources.
4. Write a one-line description after each link so a reader knows why the page matters.
5. Save the file as plain markdown named llms.txt and place it at your site root so it is served at /llms.txt.
6. Validate that the file is reachable, that every link resolves, and that the markdown renders cleanly.
7. Add the file to your deploy process so it is updated whenever you add or retire major pages.

None of these steps require special tooling. A text editor and a way to publish a static file to your root are enough. The discipline is in keeping the links honest and current.

What Is Llms-Full.Txt and Do You Need It?

The same proposal that describes llms.txt also describes a companion convention called llms-full.txt. Where llms.txt is a curated summary with links, llms-full.txt is meant to be the concatenated full text of the pages referenced -- a single large markdown document containing the actual content of your key pages rather than just pointers to them.

The theory is that a model could fetch one file and have your entire site's content in a single, token-friendly document. In practice, llms-full.txt is far heavier to generate and maintain, and it raises the same unresolved question as llms.txt: no crawler has committed to consuming it. For most marketing and web teams, a clean, well-linked llms.txt is sufficient. Treat llms-full.txt as optional, and only consider it if you already have a reliable pipeline that can regenerate the full concatenated text whenever your content changes.

Does Llms.Txt Actually Improve AI Search Visibility?

The honest answer is that there is no evidence today that publishing llms.txt measurably improves how often your site is cited by AI systems. No major provider -- Google, OpenAI, or Anthropic -- has announced that it reads or honors llms.txt as part of ranking or citation decisions. Any claim that a specific llms.txt produced a traffic lift or a citation-rate increase should be treated as unverified.

What we can say is that the file is cheap to create and carries essentially no downside, so if you find the proposal compelling, adding one is reasonable defensive hygiene. But it should be framed as a small, optional courtesy, not as a lever that moves AI visibility. The things that actually move visibility are the fundamentals: clear, crawlable, well-structured content; accurate LLM optimization practices; and deliberate AI citations work that earns references from sources models trust. For a broader view, see our guide to LLM search optimization.

What Should You Do Instead of Relying on Llms.Txt?

If your real goal is to show up in AI search and answer engines, invest in the fundamentals that models and crawlers already use. First, make sure your content is genuinely crawlable and not buried behind scripts or login walls. Second, add structured data so the meaning of your pages is explicit. Third, build topical authority by publishing substantive, accurate content that other reputable sources link to and cite.

Think of llms.txt as a name-tag you optionally pin on. The name-tag is nice, but what gets you recognized is the substance of what you say and whether others vouch for you. Spend the bulk of your effort on content quality, information architecture, and earning real citations, and treat llms.txt as a minor, optional complement rather than a strategy.

How Do You Keep an Llms.Txt File Maintained?

A stale llms.txt is worse than none, because it points models at pages that no longer exist or misrepresents what your site has become. Maintenance is light but should be intentional. The simplest approach is to tie the file to your content workflow: whenever you publish or retire a major page, update the relevant link list.

Some teams generate llms.txt from their CMS so it stays in sync automatically. Others review it quarterly as part of a content audit. Either works. The only mistake is to write it once, forget it, and let broken links accumulate. Treat it like a sitemap-lite: low effort to keep healthy, and worth keeping honest if you are going to bother having one at all.

Key Takeaways

• llms.txt is a proposed, not official, standard for summarizing your site for LLMs at /llms.txt.
• No major AI crawler has committed to honoring it, so do not expect measurable ranking or citation gains.
• It is a low-cost, harmless courtesy file -- not a substitute for crawlable content, schema, or good architecture.
• Write it as simple markdown: H1, blockquote summary, and H2 link lists; consider llms-full.txt only if you can automate it.
• Invest in the fundamentals of LLM search optimization and real citations for actual AI visibility.

Frequently Asked Questions

Is Llms.Txt Officially Supported by Google, OpenAI, or Anthropic?

No. llms.txt is a community proposal introduced by Jeremy Howard and Answer.AI in 2024. As of now, no major AI provider has announced that it officially reads or honors the file as part of its crawling, ranking, or citation process. It remains a voluntary convention with growing publisher adoption rather than an enforced standard.

Where Exactly Does the Llms.Txt File Go on My Site?

It is placed at the root of your domain and served at the path /llms.txt, the same root location as robots.txt and sitemap.xml. It must be a plain markdown file reachable over HTTPS without authentication. The companion llms-full.txt, when used, also lives at the root and is served at /llms-full.txt with the concatenated full text of referenced pages.

Can Llms.Txt Replace Robots.Txt or Sitemap.Xml?

No. robots.txt controls crawler access to paths, and sitemap.xml helps search engines discover your URLs. llms.txt does neither; it is a curated summary for language models with no enforcement mechanism. You should keep robots.txt and sitemap.xml in place and treat llms.txt as an optional addition that does not affect crawl rules or URL discovery.

Does Publishing Llms.Txt Guarantee More AI Search Traffic?

No, and you should be skeptical of any claim that it does. There is no verified evidence that publishing llms.txt measurably increases traffic or citation rates from AI systems. The file is cheap to add and harmless, but real visibility comes from crawlable, structured, authoritative content that other trusted sources cite, not from the presence of a single summary file.