Local Waifu

A small Unity-chan companion prototype using a local LLM, and local TTS. You can speak to her and get a voiced response without sending the conversation to a cloud AI service.

System Requirements

Runs a local AI model (Gemma 3 1B) + Pocket TTS fully offline — no internet, no account, nothing sent to a server.

• GPU: Not required (CPU-OK)
• RAM: 8GB min, 16GB recommended
• OS: Windows