NextFin News — StepFun has released its Step Audio 3 series of models, which are now available on the company’s open platform.
The lineup includes StepAudio 3 Realtime for full-duplex voice interaction, StepAudio 3 ASR for speech recognition, StepAudio 3 TTS for text-to-speech, StepAudio 3 Gen for unified audio generation covering speech, sound effects and music, and StepAudio 3 Music.
The models support real-time interaction, context-aware transcription, natural speech synthesis and multi-type audio creation from natural-language prompts. They are accessible via StepFun’s domestic and international open platforms.
Explore more exclusive insights at nextfin.ai.
