Does llms.txt actually affect whether AI engines cite me?

No engine has committed publicly to reading it, so treat it as cheap insurance rather than a ranking lever. What reliably helps is the underlying work it forces: naming your canonical pages, describing each one in a sentence, and being explicit about which URL answers which question.

Where do I put the file?

Serve it at https://yourdomain.com/llms.txt as plain text. On Next.js you can generate it from a route handler so it stays in sync with your real page list instead of drifting, which is how the file on this site is built.

What is the difference between llms.txt and llms-full.txt?

llms.txt is the index: a short summary plus annotated links. llms-full.txt is the full text of that content concatenated into one file, so an assistant can ingest everything in a single fetch. Publishing both is common; the index is the one to start with.

Free tool · AI search

llms.txt Generator

Build the llms.txt file that tells ChatGPT, Perplexity, and Claude which of your pages to cite.

No sign-up. We read the page to fill this in. Nothing is stored.

In the output

What you get

A sectioned llms.txt, ready to paste
Absolute links on every entry
A one-line description per page
The H1 and summary block the format requires

Which of your pages an assistant should read first, written as a ranked, sectioned list rather than a sitemap dump.
That every link is absolute, because a relative path in llms.txt resolves against the assistant’s context and not against your domain.
That each entry carries a one-line description, which is what an assistant uses to decide whether a page is worth fetching.
That the file opens with a single H1 and a summary block, the two elements the proposed format treats as required.

Three steps

How to use the llms.txt Generator

Enter your domain

We read your homepage for the name and the one-line summary an assistant reads before anything else. Correct them to say what the company does, not what the file is.

Group your pages into sections

Docs, product, pricing, guides. Sections are how an assistant decides which part of the file is relevant to the question it was asked.

Copy the output to /llms.txt

It belongs at the root of your domain, beside robots.txt, and it is a plain text file with no build step.

Why it matters

Why publish an llms.txt at all?

Assistants pick, they do not crawl everything.

A model answering a question fetches a handful of pages. This file is how you say which handful.

Your sitemap is the wrong list.

A sitemap enumerates what exists. An llms.txt nominates what is worth reading, which is usually a small fraction of it.

Descriptions do the deciding.

An assistant reads the one-line description before it fetches the page. A bare link gets skipped in favour of one that explains itself.

It costs one file.

No build step, no plugin, no schema. A text file at your root, edited when your best pages change.

After you run it

What to do with the result

An llms.txt is a request, not an instruction: it tells an assistant which pages you consider authoritative, and the assistant decides whether to honour it. That makes it worth writing well and worth revisiting, because the pages you most want cited change as you publish. Veritas tracks which of your pages actually get cited across ChatGPT, Perplexity, Gemini and AI Overviews, so the list stops being a guess.

AI Search Intelligence

Frequently asked questions

llms.txt is a markdown file at the root of your domain that tells large language models what your site is about and which pages are worth reading. It plays the role robots.txt plays for search crawlers: a single, predictable place to state what matters. It is a community proposal rather than a ratified standard, and support varies by engine.

Same cluster

More free tools

robots.txt Generator

Write a robots.txt that lets answer engines in and keeps training crawlers out, in two clicks.

AI Crawler Checker

See exactly which AI crawlers you allow, block, or never mentioned, from just your domain.

FAQ Schema Generator

Turn a page’s questions and answers into valid FAQPage JSON-LD.

See where you actually stand.

A free tool checks one thing once. Veritas watches all of it, continuously, across every engine that decides whether you get cited.

GOOGLE · CHATGPT · PERPLEXITY · GEMINI · AI OVERVIEWS