For AI & developers
Your AI already knows how.
AINP — the open format Rho speaks — is CC BY 4.0 and built for language models. Describe what you want in plain language; your LLM writes the document.
“Make an explorable that shows how a bell curve changes as I drag its mean and spread.”
```interact
slider mean range=-5..5 step=0.5 default=0 label="Mean (μ)"
slider sd range=0.5..4 step=0.1 default=1.5 label="Spread (σ)"
template vega-lite: { …a bell curve of the normal density… }
```↑ the exact block behind the bell curve on the examples page.
And it goes deeper than generation: the desktop app is AI-native end-to-end — every feature is exposed through a local agent API, so your assistant doesn't just write documents, it can open, edit, organize, and publish them for you.
Rho AI is that idea shipped inside the app: it tends your library, searches the web and shows its sources, converts Word and PDF files, and turns a table into a report with a chart. Every signed-in account gets 10 requests to try it — after that it comes with every paid plan, with a larger allowance at each step up. We run the model; there is no API key to set up.
On the web, rho.md/ai is the same Rho AI pointed at a different library: instead of your own notes it answers from everything published on Rho Cloud, and links back to each document it used. Sign in with a free account to try it.