Model Overview
Alibaba · Active Model
- Long-Context
- API Available
- Production Ready
Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a "thinking" mode for complex reasoning, math, and...
- Developer
- Alibaba
- Release Date
- April 28, 2025
- Context Window
- 131,072 tokens≈ 175 words
- API Access
- Publicly AvailableIntegrate via official API
- Input Cost
- $0.46per million tokens
- Output Cost
- $1.82per million tokens