StepFun Launches StepAudio 3 Music: A Foundation Model for Full-Song Generation

Key Info

StepFun and ACE Studio have released StepAudio 3 Music, described as StepFun's first music generation foundation model, designed to turn a prompt and lyrics into a complete song. Official resources including a Hugging Face Space, API documentation, and a thesis are now available.

Highlights

• StepAudio 3 Music is a foundation model focused on full-song generation from textual prompts and lyrics.
• It is presented as a collaboration between StepFun and ACE Studio.
• The release includes hands-on resources: a Hugging Face Space, API docs, and a technical thesis.