StepAudio 3 TTS
StepAudio 3 TTS produces speech from text and reference audio. It models spoken details such as pauses, hesitation and changes of tone, and can stream output before a full sentence is complete.
Choose the output path
The model guide documents both the HTTP speech endpoint and a WebSocket audio endpoint. These are two interfaces to the same model.StepFun stepaudio-3-tts guide
Keep spoken text and direction separate
Practical guidance. Parentheses can act as performance directions. Review text that must include literal parentheses before synthesis. Divide long narration at meaningful sentence boundaries so each request has enough context for the intended delivery.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API, Provider product |
| Model identifiers | stepaudio-3-tts |
| Control method | Use the instruction field or inline parentheses. The voice_label field is not supported for this model.StepFun speech synthesis API |
| Request limits | Up to 1,000 input characters and 500 instruction characters in the speech API.StepFun speech synthesis API |
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 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 Gen
StepFun · StepAudio 3
StepAudio 3 Gen creates audio from role descriptions, scripts and natural-language instructions.
Launch date not confirmedPreviewAPI · Provider product