Qwen3 235B A22B
alibaba
Qwen3 235B A22B is Alibaba's flagship mixture-of-experts model with 235B total parameters and 22B active per token. Delivers frontier-level performance on coding, reasoning, and multilingual tasks at significantly lower inference cost.
- Context window
- 128,000 tokens
- Input cost
- $0.00 / 1M
- Output cost
- $0.00 / 1M
- Latency (p50)
- —