Retail · Standard · pass^2
τ-bench · 8 models
| Rank | Model | Score |
|---|---|---|
| 1 | Qwen3.5-397B-A17B | 74.42 |
| 2 | GPT-5.2 | 69.59 |
| 3 | Claude Opus 4.5 | 67.4 |
| 4 | Gemini 3 Flash | 65.94 |
| 5 | Gemini 3 Pro | 63.45 |
| 6 | GPT-5.2 | 62.1345 |
| 7 | GLM-5 | 60.38 |
| 8 | Claude Sonnet 4.5 | 57.46 |
τ-bench · 8 models
| Rank | Model | Score |
|---|---|---|
| 1 | Qwen3.5-397B-A17B | 74.42 |
| 2 | GPT-5.2 | 69.59 |
| 3 | Claude Opus 4.5 | 67.4 |
| 4 | Gemini 3 Flash | 65.94 |
| 5 | Gemini 3 Pro | 63.45 |
| 6 | GPT-5.2 | 62.1345 |
| 7 | GLM-5 | 60.38 |
| 8 | Claude Sonnet 4.5 | 57.46 |