Intelligence Index
0118.2
> 37% OF MODELS
Qwen
Qwen3 235B A22B Instruct 2507
Updated: June 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
- $0.09/M
- Output
- $0.1/M
Capabilities
TEXTWEBCODINGTHINKINGWRITING
Similarly Priced Models
| Model | Provider | Context | Input Price | Output Price |
|---|---|---|---|---|
| Qwen3 Next 80B A3B Instruct | Qwen | 262K | $0.09/M | $1.1/M |
| Qwen3 Next 80B A3B Thinking | Qwen | 262K | $0.0975/M | $0.78/M |
| Qwen3 VL 8B | Qwen | 256K | $0.08/M | $0.5/M |
| Qwen3 30B A3B Thinking 2507 | Qwen | 131K | $0.08/M | $0.4/M |
| Qwen3 32B | Qwen | 131K | $0.08/M | $0.28/M |
Performance Metrics
Coding Index
0222.1
> 45% OF MODELS
Agentic Index
0322.8
> 43% OF MODELS
Average Response Performance
- Output Speed
- 57.3 tok/s
- Time To First Token
- 2.41s
- Time To First Answer Token
- 2.41s
- End To End Response Time
- 11.13s
DATA SOURCE: Artificial Analysis