Zway.ai

Free tool

Generate an llms.txt file

llms.txt is a plain text file at the root of your site that lists your most useful pages for language models in Markdown. This builds a valid one from your page list. It is a proposed convention, not a standard, so treat it as cheap insurance rather than a ranking lever.

Last reviewed 27 August 2026

Used to turn relative paths into absolute links. No trailing slash.

One page per line: path | title | description

Same format. These go under an Optional heading that models may skip.

Result

Pages listed

7

5 core, 2 optional

llms.txt

# Acme Analytics

> Product analytics for B2B SaaS teams. Event tracking, activation reporting, and account-level usage without a data warehouse.

## Core pages

- [Acme Analytics](https://acme.com/): What the product does and who it is for
- [Pricing](https://acme.com/pricing): Plans, usage limits, and what each tier includes
- [Quickstart](https://acme.com/docs/quickstart): Install the SDK and send a first event
- [Tracking plan guide](https://acme.com/docs/tracking-plan): How to name events so they stay readable
- [B2B product analytics](https://acme.com/blog/b2b-product-analytics): Why account-level reporting differs from consumer

## Optional

- [Changelog](https://acme.com/changelog): Release notes by date
- [About](https://acme.com/about): Team, funding, and company background

Before you publish

  • Publish the file at https://acme.com/llms.txt and serve it as text/plain.
  • Keep it short. The point is a curated list, so a file that mirrors your whole sitemap defeats it.
  • Update it when your top pages change. A stale file is worse than none, because it points models at pages you no longer stand behind.

Be clear-eyed about what this is. llms.txt is a community convention documented at llmstxt.org, not a specification that any AI company has committed to honouring, and no crawler is obliged to request it. It costs ten minutes and it cannot hurt you, which is the whole case for it. If you are choosing between writing this file and making your actual pages clearer, write the pages.

How this works

What the numbers mean.

  • 01Each line is split on the pipe character into path, title, and description, then rendered as a Markdown link under a heading. Relative paths are joined to your site URL so every link in the file is absolute.
  • 02Link labels have square brackets escaped and URLs have spaces and parentheses percent-encoded, so a title containing punctuation cannot produce a broken Markdown link.
  • 03The Optional heading is part of the convention: it marks pages a model can skip when it is short of context. Anything you would be annoyed to see omitted does not belong there.

Assumptions and limits

  • This produces a file. It does not verify that anything fetches it, because there is no reliable public way to confirm that from outside your own server logs.
  • Adoption is genuinely uneven. Treat any claim that a specific assistant reads llms.txt as unproven unless that company documents it.
  • The file is a pointer, not content. Models still fetch the linked pages, so a good llms.txt in front of thin pages changes nothing.

Questions about this tool

What is llms.txt actually for?
It is a curated Markdown index at the root of your domain that tells a language model which of your pages matter and what each one covers. The argument for it is that a sitemap is machine-readable but undifferentiated, while a model working with a limited context window benefits from a short ranked list. Whether crawlers use it is a separate question from whether the idea is sound.
Is llms.txt a standard?
No. It is a proposed convention documented at llmstxt.org, and no major AI company has published a commitment to reading it. Anyone telling you it is required, or that it will get you cited, is ahead of the evidence. Publish it because it is a ten-minute job with no downside, not because you expect a measurable lift.
How is it different from robots.txt or a sitemap?
robots.txt controls access and a sitemap lists every URL for discovery. llms.txt does neither: it is an editorial summary that says here are the pages worth reading and here is what each contains. The three are complementary, and llms.txt is the only one of the three that is optional in every sense.
How many pages should I list?
Enough to describe what you do and no more. Ten to thirty links is a reasonable range for most companies: product, pricing, documentation entry points, and the few pieces of writing that genuinely explain your category. If you find yourself listing every blog post, you have rebuilt your sitemap and lost the point.

This tool is free and there is nothing to sign up for. If you would rather have the work done than calculate it, that is what Zway does.