# What is llms.txt, and does your company website need one in 2026?

> llms.txt is a Markdown file at the root of a website, written for large language models. Where it came from, how to format it, which platforms read it as of 2026 an…

- Page: https://chingeo.com/en/insights/llms-txt-guide
- Updated: 2026-10-08
- Source: ChinGEO (云端智荐)
- 中文版: https://chingeo.com/insights/llms-txt-guide.md

Author: [Huang Zheng](https://chingeo.com/en/about), Founder · Systems architect · Published 2026-10-08

**llms.txt is a Markdown file at the root of a website (/llms.txt) that uses a short summary and lists of links to tell language models what the site is and where its most important content lives. It was proposed by Jeremy Howard of Answer.AI in September 2024 and is still a community proposal rather than a formal standard; no major AI platform has publicly confirmed reading it when answering. It is cheap to add, so it is worth doing, but it cannot replace server-side rendering, structured data or third-party sources.**

## The problem it solves

Web pages are written for people: navigation, scripts, ads and styling take up most of the bytes, and the useful information is buried. The idea behind llms.txt is to put a concise Markdown file at the root of the site, so language models (especially coding assistants and agents that read a site on the fly) can find the core content quickly.

## Format

Following the proposal at llmstxt.org, the file is structured like this:

1. One H1 heading with the name of the site or project (the only required part).
2. A blockquote with a one- or two-sentence summary.
3. Optional paragraphs with more detail.
4. Link lists grouped under H2 headings, one per line as "- [Page name](url): one-line note".
5. A section called Optional for secondary content that a reader can skip when context is short.

Many sites also publish llms-full.txt, which concatenates the text of the core pages into one file. This is a common practice, but it is not part of the proposal.

## Who reads it today

- **Coding assistants and agents**: some developer tools read llms.txt on documentation sites, which is its most common use.
- **Search and AI platforms**: no major platform has publicly confirmed reading llms.txt when generating answers. According to several SEO publications, Google’s search team has said it does not use the file, and it does not affect Google rankings.
- **In short**: treat llms.txt as a low-cost extra, not a core GEO technique.

## Where it sits in a GEO plan

| Priority | Work | Why |
|---|---|---|
| Highest | Render body text on the server and allow AI crawlers in robots.txt | If pages cannot be read, nothing else matters |
| High | A separate URL, title, description and structured data for every page | Lets search engines and AI understand what each page is about |
| High | Third-party sources and consistent entity information | AI trusts independent sources more |
| Extra | llms.txt / llms-full.txt | Cheap, and helpful for some tools |

## How to write it

- List only the pages that matter: services, cases, products or courses, FAQs and contact details.
- Give every link a one-line note on what question the page answers.
- Keep it in sync with the site, ideally generated automatically from the site’s data so it never goes stale.
- Do not include anything that is not on the site, and do not stuff keywords.

This site’s [llms.txt](https://chingeo.com/llms.txt) and [llms-full.txt](https://chingeo.com/llms-full.txt) are generated automatically from the site’s data and can serve as an example.

## Frequently asked questions

### How is llms.txt different from robots.txt?

robots.txt tells crawlers which addresses they may or may not fetch, and it is widely respected. llms.txt is a content guide for language models in a proposed format, and it does not control crawl permissions.

### With llms.txt, do we still need a sitemap?

Yes. A sitemap is the standard way search engines discover pages and every major search engine supports it; llms.txt cannot replace it.

### Will llms.txt improve Google rankings?

No. According to public reports, Google’s search team has said it does not use llms.txt.

---

ChinGEO (云端智荐): GEO specialist for Chinese B2B exporters: get your brand recommended by ChatGPT, Perplexity and Gemini, backed by websites, SEO and AI customer service. The site does not publish prices; costs are discussed through the project planner and a diagnosis call.

- Free AI visibility audit: https://chingeo.com/en/audit
- Site guide for LLMs: https://chingeo.com/llms.txt
- Southwest China Account Manager: Zhou Rongyue, +86 13880491401
- North China Regional Lead: Cui Yue, +86 13811954582
