Reachy Mini AI Companion

"Hey Reachy" wake-word conversational AI powered by Gemini

A voice-activated AI companion for Reachy Mini. Uses Gemini 2.5 Flash native audio for real-time voice conversations with robot expressions, gestures, and automatic session reconnection.

Features

Voice conversations — bidirectional audio via Gemini Live API
Wake word detection — "Hey Reachy" using Edge Impulse
Robot expressions — nods, shakes, looks, and gestures via function calling
Session reconnection — automatic reconnect with context preservation
Cost tracking — per-session cost estimation with daily budget limits
Personality profiles — customizable system prompts per persona
Simulate mode — web dashboard with WebRTC audio for development

Installation

From the Reachy Mini Dashboard

Install directly from the app store on your Reachy Mini dashboard with one click.

Manual Install

SSH into the robot and run:

curl -fsSL https://raw.githubusercontent.com/awakia/reachy-video-chat/main/deploy/install.sh | bash

Or from source (for development):

git clone https://github.com/awakia/reachy-video-chat.git
cd reachy-video-chat
pip install -e ".[dev]"

Features

Installation

From the Reachy Mini Dashboard

Manual Install

Links