StepAudio 2.5 TTS
StepAudio 2.5 TTS uses overall context and inline directions to control a voice performance. Its guide documents zero-shot voice cloning from a short reference recording.
Set the scene before local direction
The global instruction establishes the delivery for the whole passage. Inline instructions adjust the emotion, breath or pause at a particular point.StepFun stepaudio-2.5-tts guide
Keep instructions short and consistent
Practical guidance. A passage can contain conflicting directions. Use the global field for a stable character and reserve inline directions for genuine changes. Preview the full passage so a local instruction does not disrupt the next sentence.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API, Provider product |
| Model identifiers | stepaudio-2.5-tts |
| Input limits | Up to 1,000 text characters and 200 global instruction characters.StepFun stepaudio-2.5-tts guide |
| Inline direction | Text inside parentheses acts as a direction and is not normally read aloud.StepFun stepaudio-2.5-tts guide |
| Interfaces | HTTP synthesis and WebSocket streaming use the same model.StepFun stepaudio-2.5-tts guide |
Related models
- Step TTS 2
StepFun · Step TTS
A speech-synthesis model that uses next-token prediction without separate speaker or emotion embeddings.
Launch date not confirmedAPI - StepAudio 2.5 Realtime
StepFun · StepAudio 2.5
StepAudio 2.5 Realtime accepts and returns speech over WebSocket.
Launch date not confirmedAPI · Provider product - StepAudio 3 TTS
StepFun · StepAudio 3
StepAudio 3 TTS produces speech from text and reference audio.
Launch date not confirmedAPI · Provider product