Enter your site name, a short summary, and links to your best pages. The tool writes the Markdown for you and shows its size and estimated token count, so you can copy it or download it as llms.txt.
Basic info
Characters
16
Size
0.0 KB
Est. tokens
3
Checks (0 links)
- Add a site or project name (the H1).
- A summary (blockquote) is strongly recommended.
- There are no complete links yet. Each link needs both a title and a URL.
Your llms.txt
Untitled site
Save the file as llms.txt and upload it to the root of your site (https://yourdomain.com/llms.txt). Make sure none of the pages you list are blocked in robots.txt.
What llms.txt is and what it does
llms.txt is a plain Markdown file placed at the root of a website. It was proposed in September 2024 by Jeremy Howard, co-founder of Answer.AI and fast.ai. It gives AI agents and chatbots a short summary of your site and a curated list of links, so they do not have to dig through menus, ads, and page clutter to find what matters.
This generator builds that file from a form. You enter a site name, a summary, and sections of links with one-line descriptions. The tool assembles the Markdown, shows a preview, counts characters, bytes, and estimated tokens, and lets you copy the text or download it as llms.txt.
Keep in mind that llms.txt is a community convention, not an official standard from any one organization. Adding it costs little, but it does not promise a spot in any search result or AI answer.
Typical uses: launches, docs sites, and audits
Adding the file to a site for the first time
If you are not sure what goes in the file, the form walks you through the structure: an H1 name, a summary, sections, and links.
Pointing AI tools to your docs, pricing, and API pages
Group the pages you want AI agents to read into sections such as Docs and Pricing, and move lower-priority links into an Optional section.
Checking whether an existing file has grown too large
Paste in your links and watch the character count, file size, and estimated tokens to see if the file is still short enough to be useful.
Starting from a template for a new project
Load a blog, SaaS, online store, or portfolio sample, then replace the placeholder text and example.com URLs with your own.
Fill in the form and publish the file in 4 steps
- Enter your site or project name and a summary of one to three sentences. Add extra details if the summary leaves something out.
- Add a section and give it a name, such as Docs, Pricing, or Blog.
- Inside each section, enter a page title, its URL, and an optional one-line description for every link.
- Review the character count, estimated tokens, and checks. Then copy the Markdown or download it, and upload it to your site root as llms.txt.
Note: The Markdown is built in your browser as you type. The tool does not send your input to a server.
Limits to know before you publish
- llms.txt is a community convention, not an official standard. Some AI crawlers may never request it.
- Adding the file does not guarantee exposure or citations in AI search results or chatbot answers.
- List the pages you most want an AI to read. The file is a short guide, not a replacement for your sitemap.
- Links need both a title and a URL. Incomplete links and sections with no complete links are left out of the output.
- The 5 KB, 3,000-token, and 60-link warnings are this tool's own guidelines. The token count is an estimate, not what a specific model would report.
File structure: H1, blockquote, H2 sections, and link lists
The elements that make up the file, following the convention published at llmstxt.org.
| Element | Syntax | Role |
|---|---|---|
| Site name (H1) | # Site name | Required. The name an AI uses to identify the site or project. |
| Summary (blockquote) | > Summary text | Strongly recommended. A short statement that still helps if only the top of the file is read. |
| Extra details | Plain paragraph | Optional. Context such as audience or scope. |
| Section (H2) | ## Section name | A heading that groups related links. |
| Link item | - [Title](URL): description | One page with an optional one-line description. |
| Optional section | ## Optional | By convention, links an AI can skip when it has limited room to read. |
llms.txt is a loose convention, and AI providers may read it differently. Treat this table as a typical layout, not a strict rule set.
Does a plain-text AI guide help? What to expect
A practical look at how the file works and how to treat it.
It works as a table of contents for AI, not a ranking switch
The file saves an AI from parsing navigation, ads, and layout, so it can find your main pages and a plain-language summary quickly.
At the time of writing, major AI providers have not committed to fetching or using it, and support varies by crawler. Treat it as one inexpensive, optional step rather than a way to guarantee AI visibility.
A short, curated list says more than a long one
A long list of links uses up the space an AI has to read and buries the pages that matter.
Keep only the pages you want an AI to read first in each section, and move secondary pages to an Optional section.
The summary is the part most likely to be read
If a tool reads only the start of the file, the blockquote right after the H1 is what it sees.
State in one to three sentences what the site offers and who it is for. A concrete sentence about your audience and product is more useful than a vague slogan.
Keep it consistent with robots.txt and sitemap.xml
If you list a URL that robots.txt blocks, or a page that no longer exists, the AI gets mixed signals.
When you update llms.txt, check it against robots.txt and sitemap.xml, and confirm that every link opens.
Common mistakes: bloated files, dead links, and blocked URLs
Expecting the file to raise your AI search rankings
Some site owners treat llms.txt as an official standard and expect AI chatbots to cite them more often once it is uploaded.
💡 See the solution▼
Listing every page, so the file turns into a sitemap
Adding links on the theory that more is more produces a long file where the key pages are hard to spot.
💡 See the solution▼
Leaving out the summary and relying on the H1 alone
A file with only a site name gives an AI that reads just the top of the file almost nothing to work with.
💡 See the solution▼
Linking to wrong, unpublished, or deleted URLs
Typos from copy and paste, draft pages, and removed pages all send AI agents to dead ends.
💡 See the solution▼
Listing pages that robots.txt blocks
Linking to a page you have told crawlers not to fetch sends contradictory instructions.
💡 See the solution▼
Questions about format, placement, and privacy
Q.What is llms.txt, and how is it different from robots.txt and sitemap.xml?
A. llms.txt is a Markdown file that gives AI agents and chatbots a short summary of your site and links to its most useful pages. robots.txt tells crawlers what they may fetch, and sitemap.xml lists your pages for search engines. llms.txt is a hand-picked list with short descriptions, written for a language model to read.
Q.Will adding llms.txt improve my visibility in AI search?
A. There is no guarantee. llms.txt is a community convention, not an official standard, and AI providers handle it differently. Some crawlers may never request the file. Treat it as an optional, low-cost way to make your site easier for AI tools to understand, not as a ranking tactic.
Q.Where do I put the llms.txt file?
A. Upload it to the root of your domain so it loads at https://example.com/llms.txt, next to robots.txt and sitemap.xml. Make sure the pages you list are not contradicted by those files.
Q.How many links should I include?
A. Only the pages you most want an AI to read. The file is meant to be a short guide, not a copy of your sitemap. This generator shows a warning when you list more than 60 links, when the file passes 5 KB, or when the estimated token count passes 3,000. Those are this tool's own guidelines, not part of the llms.txt format.
Q.What does the Optional section do?
A. By convention, a section named Optional holds links that can be skipped when an AI has limited room to read. Use it for secondary pages such as an about page or case studies. It is not required.
Q.Is my site information sent anywhere?
A. No. The Markdown, the character count, the file size, and the token estimate are all produced by JavaScript in your browser. The tool does not send the names, URLs, or descriptions you type to a server.
Q.How are the estimated tokens calculated?
A. The tool counts non-space characters and divides Latin text by 4 and Japanese or other CJK text by 1.8, then rounds the sum. It is a rough guide, not the exact count any specific model would produce.
Q.What happens to links that are missing a title or URL?
A. They are left out of the generated file. A link needs both a title and a URL to be included, and a section with no complete links is skipped entirely. The checks panel tells you how many incomplete links were left out.
Your feedback
makes this tool better.
Tell us which features you'd like, what should be fixed, or which new tool you'd like to see. An engineer reads every submission and uses it to guide development.