rhino

Why we included this project

Rhino takes a different route from most voice assistants: instead of transcribing speech to text and then parsing that text, it reads intent straight from the audio. Say "a small double-shot espresso" and it returns a structured result with the relevant values already extracted. That directness is what makes it attractive for embedded work, since it runs on Arm Cortex-M, Raspberry Pi, Android, iOS, and in browsers, all on-device with no network call. You define your own command contexts and slot types in the Picovoice Console, then wire them in through SDKs for Python, Node, Flutter, React Native, and more. For teams building smart-home controls, kiosks, or other voice UIs where the command set is fixed and latency matters, the small footprint and real-time behavior are the main reasons to try it.