Yumii Character

Meet Yumii — Your AI Companion, right in your terminal.

A real-time voice AI with 6 personalities, live expressions, and a Live2D avatar. Everything runs locally for maximum privacy and performance.

V2.0.1 IS NOW LIVE
yumii — local-runtime
yumii start --personality "Genki"
Initializing local LLM engine... [OK]
Loading Live2D assets... [OK]
Yumii:"Yahoo! I'm all set up and ready to help! What's our first big mission today? ✨"

Built for Precision.

graphic_eq

Real-time Voice Chat

Experience zero-latency voice interactions with natural prosody. Yumii listens and speaks with human-like rhythm, processed entirely on your machine.

visibility

Live2D Animation

Visual expressions that react to your words. A fully integrated Live2D avatar for an immersive companion experience.

security

Privacy-First

Your data never leaves your device. No cloud logging, no data mining, no exceptions.

bolt

Lightning Fast

Optimized C++ and Python core ensuring the highest inference speeds even on consumer-grade hardware. Built with technical excellence at its core.

6 Personalities.

Choose an AI that matches your workflow or mood.

🥰
CARING

Supportive & Kind

Always there to offer encouragement and structured help for your daily tasks.

😤
TSUNDERE

Blunt & Effective

Strict, focused, and won't hesitate to tell you when your code needs fixing.

🎉
GENKI

Energetic & Bright

High-energy assistant that turns debugging sessions into a collaborative event.

🧊
KUUDERE

Cool & Logical

Calm, collected, and strictly professional. Perfect for high-stress environments.

💕
YANDERE

Devoted & Intense

Unwavering focus on your progress. She'll make sure you never miss a deadline.

🥺
DANDERE

Quiet & Precise

Soft-spoken but highly observant. Offers insights exactly when needed.

Quick Installation.

# Clone the repository
git clone https://github.com/yumii/yumii-ai.git
# Install dependencies
cd yumii-ai && ./scripts/install.sh
# Run Yumii
python main.py --personality "Caring"
CORE TECHNOLOGY
Python 3.10+
FastAPI
LangGraph
Live2D SDK
PyTorch
Whisper V3
Llama.cpp
Docker