Tongyi Fun
Explain what makes up Tongyi Fun’s enterprise audio stack, and how Fun-ASR recognition and Fun-CosyVoice synthesis cover meeting notes and support quality checks.
Tongyi Fun is Alibaba Tongyi’s enterprise audio foundation suite, pairing the Fun-ASR recognition model with Fun-CosyVoice synthesis. Fun-ASR trains on millions of hours of real speech, covers industry terms and noisy rooms, and uses retrieval to cut cross-language hallucination; Fun-CosyVoice adds multi-timbre synthesis and cross-lingual voice cloning. Both ship open source for transcription, support review and voice bots.
