Alibaba launches CosyVoice Studio, an all-in-one AI voice productivity platform

Alibaba has launched CosyVoice Studio in China, an all-in-one AI voice productivity platform powered by its own Qwen-Audio model. It includes features for voice recording, voice agents, and audio content creation. The platform is available on iOS, Android, Mac, and Windows, and is currently free to try.

The voice recording feature, CosyFlow, adds semantic understanding to transcription, filtering out filler words like 'um' and 'uh' to produce structured text. It's useful for meetings and interviews, and can identify different speakers and generate summaries. CosyAgent lets users create real-time conversational agents using natural language, aimed at customer service and telemarketing. CosyCreative offers thousands of voice options and voice cloning, allowing users to turn documents into podcasts or audiobooks. CosyAgent and CosyCreative are currently being tested with enterprise users by invitation, with a full rollout expected soon.