The short answer
/llms.txt) that gives AI models a clean, curated map of your most important content, like robots.txt or sitemap.xml but for large language models. It does not guarantee citations or override robots rules, and engine adoption is still early, but it’s low-effort, future-facing, and a clear signal of an AI-ready site. Here’s exactly what goes in one, with a real example you can copy.Key takeaways
- It's a Markdown file at /llms.txt — a curated map of your best content for LLMs.
- It is not robots.txt: it curates and explains, it doesn't gate access.
- Adoption is early and unconfirmed by major engines — treat it as a low-cost readiness signal.
- This very domain serves a live llms.txt you can inspect as a working example.
llms.txt was proposed in 2024 by Jeremy Howard of Answer.AI (llmstxt.org) as a way to give language models a concise, curated entry point to a site — because models work from limited context windows and benefit from a human-curated map rather than crawling everything. It’s part of the broader generative engine optimization toolkit.
Does llms.txt actually help SEO or get you cited?
Honestly: not yet in any confirmed, measurable way. Major engines haven’t stated they use llms.txt for ranking or citation, and you shouldn’t publish one expecting a traffic bump. Its value is different — it’s a cheap, forward-looking signal that your site is structured and AI-aware, and it gives models a clean summary if and when they do consult it. The cost is a few minutes; the downside is essentially zero. That asymmetry is why it’s worth doing even while adoption is uncertain.
What goes in an llms.txt file?
The format is simple Markdown: an H1 with your site name, an optional blockquote summary, and then H2 sections of curated links with short descriptions. Below is a trimmed example modeled on the live file this site serves — view the real one at /llms.txt.
# Black & Gold SEO
> Premium AI-powered SEO platform with 19 intelligence modules.
> Automated audits, content optimization, backlink outreach, and
> AI Overview optimization.
## Key pages
- [Features](https://www.blackandgoldseo.com/features/): All 19 tools
- [Pricing](https://www.blackandgoldseo.com/pricing/): Plans and checkout
- [Guides](https://www.blackandgoldseo.com/guides/): GEO, AEO & AI search
## Guides
- [Generative Engine Optimization](https://www.blackandgoldseo.com/guides/generative-engine-optimization/)
- [How to rank in AI Overviews](https://www.blackandgoldseo.com/guides/rank-in-ai-overviews/)How to create your llms.txt
- Create a plain-text file named
llms.txtserved at your domain root. - Start with
# Your Brandand a one- or two-line blockquote summary. - Add
## Key pagesand## Guidessections with your most important URLs and a short description each — curate, don’t dump every URL. - Use absolute URLs so the references are unambiguous.
- Keep it current as your best content changes.
llms.txt vs robots.txt vs sitemap.xml
| File | Audience | Purpose |
|---|---|---|
llms.txt | Large language models | Curate & explain your best content |
robots.txt | Crawlers (search & AI) | Allow / disallow access to paths |
sitemap.xml | Search engines | List all indexable URLs for discovery |
Sources & further reading
Keep reading
GEO · How-to
Get cited by ChatGPT
How ChatGPT, Perplexity and Copilot pick the sources they cite, which AI crawlers to allow, and the steps to become the page they quote — built from how answer engines actually work.
Pillar guide
Generative Engine Optimization
What GEO is, how AI answer engines choose what to cite, and the exact method to get your site quoted by ChatGPT, Google AI Overviews, Perplexity and Gemini — without fabricating anything.