AI search guide
llms.txt Example for SaaS, Documentation and Tool Websites
Use a practical llms.txt example for SaaS, documentation, and tool websites, with notes on what to include and avoid.
A practical llms.txt example
Use this structure as a starting point, then replace each link with canonical public resources from your own site.
# Example SaaS Public resources for understanding Example SaaS. ## Product - [Overview](https://example.com/) - [Features](https://example.com/features) - [Pricing](https://example.com/pricing) ## Documentation - [Docs](https://example.com/docs) - [API reference](https://example.com/docs/api) ## Trust and support - [Support](https://example.com/support) - [Privacy policy](https://example.com/privacy) This file is an emerging discovery aid and does not guarantee indexing, ranking, or citation.
SaaS example
A SaaS llms.txt file should prioritize the product overview, features, pricing, integrations, documentation, security pages, support, and policies. The goal is to help automated readers find the most authoritative public explanation of the product.
Documentation site example
A documentation-heavy site should prioritize getting started guides, API references, changelogs, release notes, troubleshooting, and versioned docs. Avoid old versions unless they remain publicly supported.
# Example Docs ## Start here - [Getting started](https://docs.example.com/start) - [API reference](https://docs.example.com/api) - [Changelog](https://docs.example.com/changelog)
Tool website example
A tool website should prioritize the main tool page, examples, limits, privacy notes, support, and related guides. If the tool creates reports, include documentation that explains how to interpret those reports.
# Example Tool ## Tool resources - [Run the tool](https://example.com/tool) - [Examples](https://example.com/examples) - [How to interpret results](https://example.com/guides/results)
Mistakes to avoid
Do not copy an example verbatim. The file should describe your own site. Avoid listing every URL, including private resources, linking to redirected URLs, or making unsupported claims about guaranteed LLM visibility.
Validate the example before publishing
After adapting an example, validate that the file has a readable heading, useful canonical links, plain text output, and no conflicting robots.txt policy.