# llms.txt

## Quick answer

llms.txt is a proposed plain-text file, placed at a website's root the same way robots.txt is, meant to give AI systems a short, curated list of a site's most important pages. It is a community convention, not a web standard, and it is unproven whether AI engines read or use the file.

## Why it matters

Because it is unproven that AI engines read llms.txt, it should not be treated as a way to earn AI citations. The pages an AI engine actually cites are decided by what it can crawl, read and judge relevant to a question, which llms.txt does not change on its own. Teams sometimes hear about it as a quick AI-visibility fix, and the realistic expectation is that it is unproven, not a lever worth prioritizing over the page content itself.

## How to measure it

There is nothing to measure about llms.txt specifically, since engine support is unconfirmed, so the only honest check is to keep watching whether the pages an AI engine actually cites change after adding one. The more reliable work is making sure the pages themselves, not a separate index file, directly and clearly answer the questions buyers ask. A team pressed for time is better off spending it on page content than on building or maintaining this file. If a file is added anyway, keep it short, list only the pages that best answer buyer questions, and review it whenever those pages change, so it never points at outdated material. Treat it as housekeeping with no promised payoff, and record the date it was added so any later change in citations can be compared against a clear before and after.

## Example

A documentation-heavy developer tool adds an llms.txt file after reading about it in a growth newsletter, expecting a quick lift in how often an AI engine cites their docs. A few weekly AI-visibility checks later, the citation pattern looks unchanged, while a separate round of edits that moved key setup instructions earlier on the actual doc pages does coincide with a new citation appearing. The team keeps the file in place since it cost little to add, but stops expecting it to move the number on its own.

## FAQ

### Should every brand add an llms.txt file?

It costs little to add and is unlikely to hurt, but it should not be treated as a priority over improving the actual pages an AI engine would need to read and cite.

### Is llms.txt the same as robots.txt?

No. robots.txt is a long-established standard that tells crawlers which parts of a site they may access; llms.txt is a newer convention aimed at AI systems specifically, and it is not enforced or read the way robots.txt is.

### Could llms.txt still become useful later?

It might, if AI engines eventually adopt it, but that is not established, so treating it today as housekeeping rather than a growth lever is the realistic position, worth revisiting if that changes.

## Related terms

- /glossary/ai-grounding
- /glossary/generative-engine-optimization
- /glossary/speakable

## Related

- Schema markup and AEO: /blog/schema-markup-and-aeo