funasr/demo

FunASR: Speech Recognition Demo

Industrial-grade ASR toolkit — upload audio and get transcription.

Models Available

• —SenseVoice: Multi-task (ASR + emotion + events), 5 languages, ultra-fast even on CPU
• —Paraformer: Non-autoregressive Chinese ASR with punctuation restoration

Voice Note Analyst

The Voice Note Analyst tab keeps FunASR transcription visible and editable, then optionally sends the edited transcript to Nebius Token Factory for a structured summary, key points, action items, and follow-up message. A Nebius API key is entered in the password field and used only for the selected request.

This live workflow corresponds to `Arindam200/awesome-ai-apps#250`.

Links

• —GitHub
• —Fun-ASR
• —SenseVoice
• —FunClip
• —Documentation
• —PyPI