@xcchenx345

mimo-asr transcribes audio files to text via xiaomi's free gradio api, supporting chinese and english with automatic language detection. requires only requests library and internet connectivity.

mimo-v25-tts wraps xiaomi's tts models supporting preset voices, voice design via text description, and voice cloning. covers natural language style control, director mode, emotion tags, dialects, and singing. lacks intent clarity and failure mode documentation.

mimo-tts-wav wraps xiaomi's mimo v2.5 tts engine supporting preset voices, voice design via text, and voice cloning. provides natural language style control and director mode for tone/emotion shaping. no explicit failure handling or retry logic documented.

mijia-py-api wraps xiaomi smart home device control via mijiaAPI, offering device state queries and property manipulation. lacks procedural depth, error handling, and clear decision logic for agent orchestration.

使用小米 MiMo 的多模态模型分析和理解图片、视频和音频。当用户发送图片/视频/音频附件，询问视觉内容，请求图片描述、OCR、物体检测、场景理解、视频分析或音频转录/理解时使用。