Step-Audio R1.1
StepFun · China
StepFun real-time voice reasoning model
- Release month
- Field
- Voice
- Availability
- open-weight
- Parameters
- 33B
- License
- apache-2.0 · Commercial OK
Overview
- Registered on Hugging Face January 14, 2026 · Open weights
- For spoken dialogue that reasons while speaking · Low latency · Reasoning grounded in audio
View official announcement ↗
← All models