StepAudio 2.5 Realtime
StepAudio 2.5 Realtime accepts and returns speech over WebSocket. It combines interpretation of vocal cues with the speaking controls of StepAudio 2.5 TTS.
Preview the voice before the conversation
The model guide says that a cloned voice can be auditioned with the TTS model. The resulting voice ID can then be used in Realtime.StepFun stepaudio-2.5-realtime guide
Use Chat when the answer should be text
StepAudio 2.5 Chat accepts audio or text and returns only text. It is designed to use vocal cues, such as hesitation, as well as the words in a conversation.StepFun stepaudio-2.5-chat guide
Practical guidance. Chat and Realtime are documented access paths. They are grouped here because the reviewed guides do not establish a separate trained base for each interface.
Evaluate the voice in dialogue
Practical guidance. An isolated sample does not show how a voice handles interruptions or changing emotion. Review the cloned voice in short replies, long answers and clarification questions before selecting it for an assistant.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API, Provider product |
| Model identifiers | stepaudio-2.5-realtimestepaudio-2.5-chat |
| Voice reuse | Reference-audio voice cloning uses the same quality path as StepAudio 2.5 TTS.StepFun stepaudio-2.5-realtime guide |
| Conversation setup | The interface supports persona instructions and server-side voice activity detection.StepFun stepaudio-2.5-realtime guide |
Related models
- StepAudio 2.5 TTS
StepFun · StepAudio 2.5
StepAudio 2.5 TTS uses overall context and inline directions to control a voice performance.
Launch date not confirmedAPI · Provider product - StepAudio 3 Realtime
StepFun · StepAudio 3
StepAudio 3 Realtime combines audio understanding, turn control, reasoning and tool use.
Launch date not confirmedPreviewAPI · Provider product