OpenAI: gpt-oss-120b wins on more metrics (6 of 9), but the right pick depends on what you optimize for — see the breakdown below.
| Metric | OpenAI: gpt-oss-120b | Qwen: Qwen3 235B A22B Instruct 2507 |
|---|---|---|
| Intelligence Index | 33.3✓ | 25.0 |
| Coding Index | 28.6✓ | 22.1 |
| GPQA Diamond | 78%✓ | 75% |
| Design Arena Elo | 1062 | 1103✓ |
| Speed (tokens/sec) | 699✓ | 86 |
| Latency | 149ms✓ | 287ms |
| Input price /M | $0.039✓ | $0.090 |
| Output price /M | $0.180 | $0.100✓ |
| Context window | 131K | 262K✓ |
| Capabilities | ReasoningToolsJSON | ToolsJSON |