Do I need an llms.txt file?
Short answerYou don’t need one. llms.txt is a proposed convention, not a standard, and the major AI search tools haven’t committed to using it. Google has said it doesn’t rely on it. It’s cheap and harmless to add, so it can be worth ten minutes, but it won’t replace a clear, crawlable website.
Want your site ready for AI crawlers?
I’ll check your robots file, llms.txt and whether AI tools can read your pages. Free, within two business days.
The longer answer
What llms.txt is
llms.txt is a plain Markdown file placed at yourdomain.ca/llms.txt. It was proposed in 2024 as a way to give language models a short, curated summary of a site: what it is, and links to the most important pages, sometimes with cleaner text versions.
The idea is sensible. Web pages are full of navigation, scripts and layout that a model has to strip away. A tidy summary could help it find the right content faster.
Who actually reads it
This is the uncertain part. At the time of writing, the big AI search products haven’t publicly committed to using llms.txt when deciding what to cite, and Google has said its systems don’t rely on it. Some developer tools and documentation platforms do use it, mainly so coding assistants can read technical docs.
So for a small business site, the realistic benefit today is small and speculative. That may change. Conventions like robots.txt and sitemaps took time to be adopted, and some never were.
If you add one, keep it simple
A good llms.txt is short: your business name, a one-paragraph description of what you do and where, and links to your key pages with a line about each. Keep it accurate and update it when your services change.
Don’t put anything in it that isn’t on your site, and don’t treat it as a place for keyword stuffing. An llms.txt generator can produce a clean starting file in a few minutes.
What I’d do
Nick Lovett, founder of Tavro
I add an llms.txt to the sites I build because it takes a few minutes, it’s generated from the same page list as the sitemap, and it does no harm. I tell clients plainly that it’s a bet on a convention that may or may not catch on.
The things I’d spend real effort on first are the ones that clearly matter for AI search: prerendered HTML that any crawler can read, clear answers on each page, and consistent business information across the web.
Follow-up questions
No. robots.txt tells crawlers what they may access. llms.txt is a summary meant to help AI tools understand your site. It doesn’t grant or block anything.
There’s no evidence it does on its own. Clear, crawlable content and a good reputation matter far more.
Not in any known way, as long as it’s accurate. The main cost is keeping it up to date.
Keep reading
- nick@tavro.ca
- 587-439-9994
- Calgary, Alberta