AI Audio

Aero-1-Audio Aero-1-Audio

A lightweight audio model with only 150 million parameters, supporting 15 minutes of continuous audio processing

Fun Asr Fun-ASR

Tongyi Lab’s open source end-to-end speech recognition model supports Chinese dialects and 31 languages.

AInterview AInterview

You are the guest and AI is the host, generating a podcast interview in a few minutes

Fun-AudioGen-VD Alibaba Tongyi Lab

The timbre design model launched by Alibaba Tongyi Lab supports natural language description to generate timbre, emotion and scene-based audio.