Free llms.txt Generator
Build a clean, spec-compliant llms.txt file, a curated map of your best content for AI models. Free, no signup. Everything runs in your browser. New to this? Read what llms.txt is and whether it works.
# Acme Docs > Acme is an open-source toolkit for building data pipelines. > This file points AI models to the docs that matter most. ## Core documentation - [Getting started](https://example.com/docs/start): Install and run your first project. - [API reference](https://example.com/docs/api): Every public function, with arguments. ## Guides - [Deploy to production](https://example.com/docs/deploy): Production checklist and scaling notes. ## Optional - [Changelog](https://example.com/changelog): Release history.
Save the file as llms.txt and upload it to your site root, so it resolves at yoursite.com/llms.txt (same level as robots.txt). Then check it loads as plain text.
How to use it
- Name the site or project. This becomes the single top-level heading of the file, and it is the first thing a model reads.
- Write one or two sentences of summary. Say what you do and who it is for, in plain language. This becomes the quoted block under the heading.
- Group your best links under headings. Each link takes a title, a URL and a short description. Add or remove sections and links freely.
- Copy or download, then upload the file to your site root so it loads at
yoursite.com/llms.txt.
The preview rebuilds on every keystroke, so there is no generate button and no waiting. The form starts filled with a worked example so you can see the shape before you replace it with your own.
What the finished file looks like
Here is exactly what the generator produces from the example it loads with. Yours will have the same shape with your own headings and links:
# Acme Docs
> Acme is an open-source toolkit for building data pipelines.
> This file points AI models to the docs that matter most.
## Core documentation
- [Getting started](https://example.com/docs/start): Install and run your first project.
- [API reference](https://example.com/docs/api): Every public function, with arguments.
## Guides
- [Deploy to production](https://example.com/docs/deploy): Production checklist and scaling notes.
## Optional
- [Changelog](https://example.com/changelog): Release history.What each field becomes
Every box in the form maps to one line pattern in the output. If you want to understand the file rather than just generate it, this is the whole format:
| Field in the form | Line it produces | What it is for |
|---|---|---|
| Site / project name | # Acme Docs | The one required line. The name a model reads first. |
| Summary | > Acme is an open-source toolkit… | A quoted block giving context before any link. Each new line becomes its own quoted line. |
| Section heading | ## Core documentation | Groups related links so a model can tell reference from tutorial from release notes. |
| Link title, URL, description | - [Getting started](https://…): Install and run… | One line per page. The description is what tells a model whether to open it. |
A section named Optional | ## Optional | In the llms.txt format this heading has a specific meaning: links a reader may skip when a shorter version is needed. Keep the spelling if you want that meaning. |
Two small behaviours worth knowing. A link row with an empty URL is skipped rather than written as a broken line, so you can leave a spare row lying around. And a link with no description is written as a bare link, which is valid but wastes the most useful part of the line.
What this generator does not do
- It does not crawl your site. You pick the links by hand. That is the point: llms.txt is a curated shortlist, not a second sitemap. If you paste in two hundred URLs you have built a worse sitemap, not a better llms.txt.
- It does not check that your URLs resolve. A typo goes straight into the file. Open each link once before you publish.
- It does not publish anything. There is no connection to your site. You upload the file yourself.
- It does not send your text anywhere. There is no server call and no model in the loop. Everything happens in the page you have open, which is also why it keeps nothing when you close the tab.
- It cannot make anyone read the file. That is the honest part, and it is worth its own section.
Does llms.txt actually do anything?
Google has said publicly that it ignores llms.txt. That is the clearest statement any major engine has made, and it means the file will not help you in Google AI Overviews or AI Mode. No other large assistant has confirmed that it reads llms.txt either. What the file does have going for it is that it is cheap, it is small, and it does no harm: an ignored text file at your root costs you nothing.
So treat it as ten minutes of tidying, not as a growth lever, and do not let it displace the work that actually moves the needle. We wrote the full version of this argument, with who said what and when, in our honest assessment of llms.txt. One thing people confuse constantly: robots.txt and llms.txt are unrelated. robots.txt is a real, honoured standard that controls whether crawlers may fetch you at all, and it is the one that can genuinely make you invisible. Check that one with the AI Bot Checker.
When it is worth making one
Three situations where we would actually spend the ten minutes:
- You run documentation or a developer product. This is where llms.txt started and where it makes most sense: a lot of pages, a clear hierarchy, and readers who genuinely paste your docs into an assistant. A curated index saves them guessing.
- Your best pages are buried. If the five pages you most want quoted sit four clicks deep behind a marketing site, a flat list of them at your root costs nothing and might help someone find them.
- You want the exercise more than the file. Being forced to name your ten best pages and describe each in one line is a useful audit in itself. Most teams discover they cannot describe half of them.
When we would skip it: a small site where every page is already one click from the homepage, or any situation where making the file would delay fixing a blocked crawler or a page with no date and no author on it.
Where the file goes, and how to check it worked
Upload it to the root of your domain so it resolves at yoursite.com/llms.txt, the same level as robots.txt. Then open that URL in a browser and confirm two things: the file loads, and it displays as plain text rather than downloading or rendering as HTML. If your host serves it with the wrong content type you will see the browser offer to download it, which is a sign to fix the type before you tell anyone the file exists.
Questions people ask
Is llms.txt the same as robots.txt?
No, and the difference matters. robots.txt controls access: it tells crawlers which parts of your site they may fetch, and blocking one there really does remove you from that engine. llms.txt suggests reading order and has no control over anything. One is a gate, the other is a note pinned to the gate.
How many links should I include?
Few enough that a person could read the whole file. The value is the curation. If you cannot justify why a page is on the list, leave it off.
Does the generator save my work?
No. Nothing is stored, on our side or yours, so download the file before you close the tab. Nothing you type is transmitted anywhere either.
Can I edit the file afterwards?
Yes, it is plain text. Open it in any editor and change a line. Coming back here and rebuilding it is only worth doing if the structure changed.
Do I need a Markdown copy of every page too?
Some versions of the convention suggest also publishing plain-text copies of individual pages. This generator does not do that, and given that the format is not confirmed to be read by the major engines, we would not spend the effort until that changes.
The other three tools
AI Bot Checker tells you whether the crawlers behind ChatGPT, Claude, Perplexity and Gemini can reach your site at all. AI Readiness Checker scores one page out of 100 on the signals that make it quotable. Prompt Generator builds the buyer questions to paste into the engines so you can see where you are missing. All free, all on our tools page.