Feature request: Add FunASR for voice-based AI wearable

[LauraGPT]

Note

License and capability clarification (2026-07-14): FunASR is a toolkit, not a single checkpoint. The FunASR and SenseVoice repository source code is MIT; model weights follow each model card. SenseVoiceSmall supports Chinese, Cantonese, English, Japanese, and Korean, and its weights use the linked FunASR Model Open Source License Agreement. Fun-ASR-Nano-2512 is Apache-2.0. Language coverage, punctuation, and performance depend on the selected model and runtime configuration.

Feature Request

Add FunASR as a self-hosted STT option for ADeus voice interactions.

Why

ADeus is a voice-first AI wearable. FunASR fits well:

• SenseVoice (234M): 50+ languages, 25x realtime on CPU — efficient for wearable use
• Emotion detection: Detect user mood during conversations
• OpenAI-compatible API: funasr-server serves POST /v1/audio/transcriptions
• Self-hosted: Aligns with open-source wearable philosophy
• ONNX models via Sherpa-ONNX for edge deployment

References

• FunASR — 16K+ stars
• SenseVoice — 8K+ stars