Description
StepFun StepAudio 3 Music (preview) — text-to-music from the StepAudio 3 series on POST /v1/music/generations, with the same request shape as minimax-music-v3.0: describe the style in prompt, add lyrics or set is_instrumental, and get the finished song back hex-encoded in data.audio (mp3 by default; wav, flac, opus or pcm through audio_setting.format). The call is synchronous and a song usually takes one to several minutes, so give the client a timeout of at least 600 seconds. Covers and accompaniment for an uploaded vocal are not offered. Free while StepFun's preview runs; StepFun retires the preview name when the free period ends and releases a paid version, so plan for this model ID to stop working at that point.
Capabilities
- Input modalities
- text
- Output modalities
- audio
- Tasks
- music-generation
API endpoints
POST /v1/music/generations(music-generation)