Intelligence Index
0118.4
> 27% OF MODELS
Qwen
Qwen3 235B A22B Instruct 2507
Updated: September 2026
Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...
Specifications
- Context
- 262K
- Input Modalities
- Text
- Output Modalities
- Text
- Input
- $0.0875/M
- Output
- $0.35/M
Capabilities
TEXTWEBWRITING
Similarly Priced Models
| Model | Provider | Context | Input Price | Output Price |
|---|---|---|---|---|
| Gemma 4 31B | Google | 262K | $0.09/M | $0.34/M |
| Qwen3 32B | Qwen | 131K | $0.08/M | $0.28/M |
| Gemma 3 27B | Google | 131K | $0.08/M | $0.45/M |
| Z.ai: GLM 5.3 Flash | Z-AI | 1M | $0.075/M | $0.25/M |
| ByteDance Seed: Seed 1.6 Flash | ByteDance | 262K | $0.075/M | $0.3/M |
Performance Metrics
Average Response Performance
- Output Speed
- 58 tok/s
- Time To First Token
- 2.36s
- Time To First Answer Token
- 2.36s
- End To End Response Time
- 10.99s
DATA SOURCE: Artificial Analysis