What Is llms.txt, and Why Your Blog Should Have One

As readers increasingly ask AI tools their questions, a small file is quietly becoming part of a well-run site: llms.txt. It is a plain-text file, served at the root of your domain, that gives large language models a clean, curated summary of what your site is and where its important content lives. Think of it as a friendly briefing for the machines that now read on your readers' behalf. This guide explains what it is, how it fits alongside the files you already publish, and how to add one.

What llms.txt actually is

llms.txt is a Markdown-formatted text file at /llms.txt. At its simplest it opens with your site name as a heading, a one-line summary, and a short description, followed by curated links to your most important pages with a sentence explaining each. The format is deliberately human-readable: a person and a language model should both be able to skim it and understand your site in seconds. You can view Inkwell's own llms.txt as a working example.

llms.txt vs llms-full.txt

Two files often travel together. llms.txt is the concise index: names, summaries, and links. llms-full.txt is the expanded version that inlines more of the actual content, so an engine that wants depth can read a fuller corpus in one request without crawling every page. The short file is the map; the full file is the territory. Inkwell generates both automatically, and you can see the llms-full.txt alongside the index.

How it differs from robots.txt and sitemap.xml

It is easy to assume llms.txt duplicates files you already have, but each does a different job:

• robots.txt tells crawlers what they may and may not access. It is permission, not explanation.
• sitemap.xml lists every URL for discovery. It is a machine index with no meaning attached.
• llms.txt explains and prioritises. It says what matters and why, in language a model can lift directly.

In other words, robots.txt gates access, sitemap.xml enumerates pages, and llms.txt curates understanding. They complement each other rather than compete.

Why your blog should publish one

An llms.txt will not force an engine to cite you, and adoption is still early, so treat it as one helpful signal rather than a magic switch. But the upside is real and the cost is tiny. It reduces the chance a model misreads your site, it surfaces your best pages instead of leaving them buried, and it states your key facts, name, purpose, and links, in one authoritative place, which helps with the entity clarity that answer engines rely on. It is a natural companion to the wider practice of answer engine optimization.

robots.txt controls access and sitemap.xml lists pages, but llms.txt is the first file that tells an AI what your site actually means.

How to publish an llms.txt

You can write one by hand: create a text file, add your site name as a top-level heading, a blockquote summary, a short description, then sections of curated links with one-line notes. Keep it current as your best content changes, and serve it at your domain root so tools can find it at the conventional path. If maintaining it by hand sounds fragile, that is exactly the kind of thing a platform should handle for you.

Inkwell generates a per-site llms.txt and llms-full.txt from your published content automatically, so they stay in sync as you write, with no file to hand-edit or forget. The details are in the documentation.

• #AEO
• #AI Search
• #Guides
• #Standards