Generator
Everything you type stays on this device. Only the site name is required in an llms.txt file.
This page blocks every outside connection, so it cannot download a sitemap for you. Open your sitemap in another tab, copy its text and paste it into the box below.
Your list of links
Each row becomes a line in the file with the title, the page address and a short description. The section is the heading the link is listed under: Docs, Application Programming Interface, Blog, Legal, Pages, or Optional for secondary links that an assistant can skip when space is short.
Your file
The text below updates as you fill in the boxes.
How it works
- Name the site. The tool writes the site name as the main heading, which the layout requires. Then it adds the short description as a quote and the extra details as a paragraph, with no further headings.
- Choose the links. Add rows with a title, a page address and a one-line description, and pick a section: Docs, Application Programming Interface, Blog, Legal, Pages or Optional. The website address fills in the start of any page address you type in part.
- Bring in a sitemap if you have one. Paste the sitemap text and the tool reads every page address in it, whether the file is a list of pages or a list of other sitemaps. It works out a title and a section from each address, and you can edit or delete any row. The Fetch sitemap button tries to download the file, but this page blocks outside connections, so it shows a note and keeps the paste box open.
- Two files.
llms.txtstays short, with all the links you chose and a limited number of sitemap pages. Tick the box forllms-full.txtto get longer notes and a higher limit. Both are Markdown lists of the form- [Name](url): notes. - Copy or download. You publish the file yourself. This page never publishes it.
The limits below come from this tool, not from the llmstxt.org layout. llms.txt includes all the links you chose plus up to 40 pages taken from a sitemap. llms-full.txt includes all the links you chose plus up to 200 pages taken from a sitemap. One sitemap import adds at most 150 rows. Remove thin pages before you publish, because a sitemap is not a substitute for a hand-picked map.
Questions and answers
Who made this?
UtilForge makes it. UtilForge is a website of free cybersecurity tools and small calculators. The About page explains why the tools were made, and the Contact page is the way to reach us.
Is this a scam site?
Here are facts you can check instead of taking on trust. The tool itself does not ask for payment, an account or card details. The box near the bottom of the page offers the five website tools as one free file that opens offline. The page also carries a security setting that stops the browser letting it send anything off your device, which is why the Fetch sitemap button cannot download a file. To test it, open the page, switch your internet off and fill in the boxes: the file text keeps updating, because it is made on your device. You do not have to take UtilForge's word for any of this.
What happens when I close the page?
The page does not save anything in your browser. It uses no cookies and no stored settings, and the page address does not change to hold what you type, so there is no share link. When you close the page, what you typed and the rows you made are not kept by it. Only what you chose to take away stays: a file you downloaded stays in your downloads folder, and text you copied stays wherever you pasted it. The file stays on your device, so UtilForge holds no copy.
Does it cost anything, and do I need to sign in?
The tool is free and you can start straight away. Nothing on the page asks for a card, and nothing on it costs money. The five website tools are also available as one free file that opens offline.
What is an llms.txt file, and where does it go?
An llms.txt file is a plain text file, written in Markdown (headings and lists made with symbols such as # and -), that gives artificial intelligence assistants a short, hand-picked list of the most useful pages on a website. The idea was proposed at llmstxt.org. The usual place for it is the top level of your site, at https://your-domain/llms.txt. A file can also sit in a folder, such as /docs/llms.txt, and then covers the pages under that folder. Serve it as text/plain or text/markdown. Artificial intelligence companies use it unevenly, and nothing promises more visits, a higher ranking or more citations.
Can I make an llms.txt file from my sitemap?
Partly. A sitemap is the list of every page on a site, saved in a file format called Extensible Markup Language. Open your sitemap in another browser tab, copy its text and paste it into the box on this page. The tool reads each page address in it (the loc entries of a sitemap or a sitemap index), works out a title and a section from the address, and adds rows you can edit. The Fetch sitemap button tries to download the file for you, but this page is set to block every outside connection, so it cannot, and it says so. The llms.txt file takes up to 40 sitemap pages and the llms-full.txt file up to 200, and one import adds at most 150 rows. A sitemap lists everything, while llms.txt should be hand-picked, so delete the rows for thin or less important pages before you publish the file.
Is llms.txt the same as robots.txt?
No. A robots.txt file says which pages crawlers may visit. An llms.txt file is an optional list of important page addresses, for tools that choose to read it. Community explainers and commentary from Google treat them as different jobs.
Will publishing llms.txt get me cited in ChatGPT or artificial intelligence Overviews?
There is no reliable public guarantee. Treat llms.txt as a helpful file for assistants that read it, not as a switch for rankings or citations. Use of it is uneven.
Do I still need a sitemap if I have llms.txt?
Yes, for ordinary search. An Extensible Markup Language sitemap, sent through Search Console or Bing Webmaster, is still the standard way for search engines to find your pages. llms.txt does not replace it.
Should llms.txt list every page on my site?
Community advice usually favours hand-picked pages of real value (documents, products, key guides) over a dump of everything a crawler could find. Clarity matters more than volume.