Updated October 4, 2026. Originally published in November 2024, this article covers a proposal for helping AI agents find website information through llms.txt. The documented August 2026 version 2 update clarifies the format; it remains a proposal rather than a guarantee that every AI service will use it.
What Is llms.txt?
llms.txt, proposed by Jeremy Howard in September 2024, is a curated Markdown entry point: a short site overview with links to selected resources.
Why Do We Need llms.txt?
The intended workflow is to read the overview, then follow relevant links. Publishing the file alone does not establish that a particular assistant retrieves it or answers accurately.
The retained format illustration below shows a title, optional description and details, section links and an Optional list. Its link_url addresses are placeholders, not a working configuration.
How Does llms.txt Work?
A file can sit at the root or a subpath. Version 2 defines coverage beneath that path and selects the most specific file. The H1 project name is required; the description, details and link sections are optional.
Illustrative template; replace the placeholder links before publishing:
# Project Name
> Brief summary or description.
Optional details or notes.
## Section Title
- [Link Title](https://link_url): Optional link details.
## Optional
- [Additional Link](https://link_url)
Version 2 permits Markdown URLs that append .md or replace the HTML extension. It also specifies alternate and describedby link relations for discovery. An Optional section still marks secondary links; the old context-expansion machinery is no longer part of the proposal.
An Example from FastHTML
The FastHTML project provides a concrete public example. These addresses served the named documents when checked on October 4, 2026:
Try It on Your Site
Start with a small set of useful pages, then check the assistant you actually intend to use.
- Choose maintained documentation or policy pages rather than copying every page.
- Publish genuine Markdown links and keep them current.
- Ask your intended assistant to start from the file; check the retrieved pages and resulting answers.
- Separate Kingy reading: ChatLLM documents and presentations. This is not an llms.txt integration example.
Working with Existing Web Standards
The proposal treats llms.txt as complementary to existing web files:
- robots.txt expresses crawler-access rules.
- A sitemap lists URLs for discovery.
A curated file does not prove an AI service uses it, or settle copyright, licensing or traffic outcomes. The wordmark images are editorial illustrations.
Sources
- Current proposal and version 2 changes.
- Jeremy Howard’s proposal.
- FastHTML: example project.
- Original related reporting: The Decoder’s article.
The Kingy Brief
Get the free AI price sheet.
26 AI plans and models on one page, from official pricing pages, updated every month. Free when you join The Kingy Brief: AI changes, original tests and one practical thing to try, Fridays at 09:00 Vancouver time.
Free · Double opt-in · Unsubscribe anytime
Sign up for The Kingy Brief on Beehiiv.
Signup help and newsletter schedule
Signup form provided by Beehiiv. After submitting, check your inbox for "Confirm your subscription to The Kingy Brief" and open its confirmation link. Your price sheet download is on the page that opens. Check Spam or Promotions if you cannot find it.
Fridays at 09:00 Vancouver time: source-checked AI changes, original tests and one practical thing to try. We skip a week when there is not enough verified material. Free. Unsubscribe anytime.