Step-Audio R1.5
Step-Audio R1.5 adds human-feedback training for natural dialogue while reasoning and speech proceed together.
Separate the hosted update from published weights
R1.5 replaced the hosted R1.1 model. R1.1 weights remain a separate release.StepFun audio model overview
Evaluate the answer and the wait together
Practical guidance. For spoken reasoning, the useful result includes both correctness and conversational timing. Listen for whether early speech commits to an answer before the reasoning or tool result is ready.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API, Provider product |
| Model identifiers | step-audio-r1.5 |
| Access | The Realtime API and billing catalog list step-audio-r1.5.StepFun Realtime API model supportStepFun billing catalog and 9 November 2026 audio change |
Related models
- Step-Audio 2
StepFun · StepAudio 2
Step-Audio 2 understands speech, music, environmental sound and emotion, with tools and search.
Launch date not confirmedAPI · Provider product - Step-Audio R1.1
StepFun · Step-Audio R
An audio model that separates higher-level reasoning from speech generation through a dual-brain design.
Launch date not confirmedDownloadable - StepAudio 3 Realtime
StepFun · StepAudio 3
StepAudio 3 Realtime combines audio understanding, turn control, reasoning and tool use.
Launch date not confirmedPreviewAPI · Provider product