Feature: Add FunASR/SenseVoice as self-hosted STT

[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.

Hi! Great AI meeting assistant!

Would you consider FunASR or SenseVoice?

Why?

• Ultra-fast — SenseVoice-Small: ~70ms/10s
• Speaker diarization — CAM++ built-in
• Built-in punctuation — Clean meeting transcripts
• Self-hosted — No API costs
• 50+ languages, Apache 2.0

References

• FunASR: https://github.com/modelscope/FunASR (17K+ stars)
• SenseVoice: https://github.com/FunAudioLLM/SenseVoice (8K+ stars)

[LauraGPT]

Hi! FunASR/SenseVoice supports real-time streaming via WebSocket and is fully self-hosted. Would be a natural fit for joinly. Let me know if I can help! 🙏