llms.txt
llms.txt is a proposed convention for a Markdown file at a site's root (/llms.txt) that gives AI models and agents a short summary of the site and links to its most useful pages.
- Also called
- llms txt, LLMs.txt file, llms-txt, LLM text file
- Category
- SEO and AEO
- Updated
The proposal was published by Jeremy Howard on 3 September 2024, and a second version followed in August 2026. The reasoning is simple: web pages are built for people, with navigation, scripts and layout that waste an AI model’s limited context. A plain Markdown file gives a model the key facts and the right links instead.
The format is light. An H1 with the site or project name is the only required part. After it comes a blockquote with a short summary, then optional H2 sections that list links with a one-line note each, and an “Optional” section for links an agent can skip when it needs a shorter context. The proposal also suggests serving Markdown versions of key pages. Documentation platforms often generate the file automatically, and OpenAI, Anthropic and Google publish one for their own developer docs.
Why it matters for a SaaS website
Be clear about what it is: a proposed convention, not a standard that search or answer engines have committed to. Google’s guidance says no AI text files or special markup are needed to appear in AI Overviews or AI Mode. Chrome’s Lighthouse now has an agentic browsing audit that checks for the file, but marks a missing one as not applicable because providing it is optional.
Where it helps is with AI agents and coding tools that read docs on request, and as a cheap, low-risk summary of who you are and which pages matter. For a SaaS company with public docs or an API, that use is real. For a marketing site, it is a small addition, not an AEO strategy.
How DesignMe handles it
Our own site serves a generated /llms.txt. It is built from the same data as the pages, so it never drifts: a short description of DesignMe, then sections for services, case studies, platform comparisons and insights, each with a one-line summary. Only indexable pages are listed; drafts stay out. For clients we treat it as an optional extra and spend most of the effort on what answer engines demonstrably use: crawlable HTML, answer-first copy, specific facts and schema markup.
Example
Say a developer tools company has 300 docs pages. Its /llms.txt opens with the product name and a two-sentence summary, then lists Quickstart, Authentication, API reference and Pricing, with SDK changelogs in the Optional section. When a developer asks a coding agent how to authenticate, the agent can go straight to the right page instead of guessing from an old blog post.
Common mistakes
- Treating it as an SEO lever. No major search engine has said it uses the file for ranking.
- Writing it by hand once. A static file goes stale. Generate it from your content source.
- Listing everything. It is a curated map, not a second sitemap.
- Contradicting the site. Prices, names and claims in the file must match the pages.
Questions
- Does llms.txt help a site get cited by ChatGPT or Google AI Overviews?
- There is no evidence that it does today. Google says you don't need AI text files or special markup to appear in its AI features, and the major AI providers have not said their answer engines use a site's llms.txt to choose sources. Citations still come from pages that rank, load as HTML and answer the question clearly.
- Is llms.txt the same as robots.txt?
- No. robots.txt tells crawlers what they may and may not access. llms.txt grants or blocks nothing: it is a curated reading list in Markdown that an AI tool or agent can open when it needs to understand a site quickly. A site can have both, and blocking a bot in robots.txt is not undone by llms.txt.
Sources
- llmstxt.org: The /llms.txt file (proposal) (opens in a new tab)llmstxt.org
- Google Search Central: AI features and your website (opens in a new tab)developers.google.com
- Chrome for Developers: Lighthouse llms.txt audit (opens in a new tab)developer.chrome.com
Press Escape to close.