Local AI

A complete local AI voice interaction pipeline:

Speech Recognition (STT)
SenseVoice model, supporting real-time transcription in Chinese, English, Japanese, and Korean.

Large Language Model (LLM)
Qwen2.5-7B-Instruct, local inference with GPU acceleration support.

Text-to-Speech (TTS)
Kokoro TTS engine, 10+ languages, 100+ voices, natural and fluent.

Live2D Presentation

Dynamic character models with eye tracking and lip sync. Supports built-in Live2D models or static images (PNG/GIF). Transparent window blends seamlessly with your desktop.

Privacy & Offline

• All AI computation runs locally
  
• No internet connection required
  
• Conversation data never uploaded to any server
  
• No API keys or third-party services needed

Ready Out of the Box

Install and start using immediately. No Python environment, model downloads, or technical knowledge required.

30 Interface Languages

简体中文 | English | 日本語 | 한국어 | Deutsch | Français | Español | Português | Русский | العربية | ไทย | Tiếng Việt and more

Legal Notice

This content uses sample data owned and copyrighted by Live2D Inc. The sample data are utilized in accordance with terms and conditions set by Live2D Inc. This content itself is created at the author's sole discretion.