It is most useful for documentation and developer products, where coding assistants read docs on request. For a local business or a store, the checks in the GEO checklist matter far more.
The llms.txt Format
The proposal at llmstxt.org defines a simple structure: an H1 with the site name, a short summary in a blockquote, optional notes, then H2 sections with lists of links, each with an optional description.
# Northwind Outdoors
> Outdoor gear shop in Portland, Oregon, selling tents, sleeping bags and backpacks online since 2014.
## Guides
- [How to choose a sleeping bag](https://northwind-outdoors.com/guides/how-to-choose-a-sleeping-bag): temperature ratings, fill types, sizes
- [Tent footprint vs tarp](https://northwind-outdoors.com/guides/tent-footprint-vs-tarp): when you need each
## Shop
- [Tents](https://northwind-outdoors.com/tents/): 2 to 6 person tents with specs and prices
## Optional
- [About us](https://northwind-outdoors.com/about/)The example is for a made-up store. The proposal also suggests offering Markdown versions of pages at the same URL plus .md. That part is optional too.
Does llms.txt Help You Show Up in AI Answers?
| Engine | What it has said | What decides inclusion |
|---|---|---|
| Google AI Overviews and AI Mode | "No additional requirements" and no "special optimizations" needed | Being indexed and eligible for a snippet |
| ChatGPT search | Its crawler docs do not mention llms.txt | OAI-SearchBot access and its search results |
| Perplexity | Its crawler docs do not mention llms.txt | PerplexityBot access and its search results |
| Claude | Its crawler docs do not mention llms.txt | Claude-SearchBot access and its search results |
Sources: Google, OpenAI, Perplexity, Anthropic. AI search bots read your normal HTML pages. Allowing them in robots.txt matters far more than a summary file.
When llms.txt Is Worth Adding
- Developer tools and APIs: coding assistants fetch docs on request, and a clean index of your docs helps them find the right page.
- Large documentation sites: a curated list of the pages that matter.
- Any site, as a cheap extra: it takes 20 minutes and does no harm, once the checks that do matter are done.
How to Add an llms.txt File
- Write the file in the format above. Keep the summary to one or two factual sentences.
- Link only pages you want AI tools to read first: key guides, products, pricing, About.
- Upload it so it loads at
yoursite.com/llms.txtas plain text. - Update it when those pages change.
On hosted platforms you may not be able to place a file at the root. Squarespace sites: see Squarespace AI crawlers and llms.txt. On other platforms, check whether your plan lets you serve a file at the root before you spend time on it.
llms.txt vs robots.txt
robots.txt tells bots what they may crawl, and OpenAI, Anthropic and Perplexity say their search crawlers respect it. llms.txt suggests what to read, and nothing obliges any tool to use it. If you only do one, check robots.txt: blocking OAI-SearchBot, PerplexityBot or Claude-SearchBot keeps you out of those engines' answers. See AI crawlers and robots.txt.
How to Use This with an AI Assistant
Connect SEOmatic at https://seomatic.ai/mcp and ask: "Sign me in with connect_seomatic. Check my robots.txt for OAI-SearchBot, PerplexityBot and Claude-SearchBot, then use gsc_performance to list my 10 most-clicked pages and draft an llms.txt file that links them with one-line descriptions." Drafting the file changes nothing on your site.
Let the Agents Do the GEO Work
SEOmatic tracks whether ChatGPT, Claude and Google AI Overviews name your business for the questions you choose, then its agents work on the pages those answers should quote:
- See which questions name you and which name a competitor
- Rewrite key pages to open with clear, quotable answers
- Add FAQ blocks and schema that match the page
Each change can wait for your approval, one click undoes it, and the agents check it against how the rest of your site did. What they can change depends on your platform, and you see the full list before anything runs.
14-Day Trial for $1, credited to your first payment.


