wide Research
7 models evaluated
|
| 1 | Kimi K2.6 | Moonshot AI | Open | 80.8% | |
| 2 | Claude Opus 4.5 | Anthropic | Closed | 76.4% | |
| 3 | Qwen3.6 Plus | Alibaba | Closed | 74.3% | |
| 4 | Qwen3.5 397B | Alibaba | Open | 74.0% | |
| 5 | Kimi K2.5 | Moonshot AI | Open | 72.7% | |
| 6 | GLM-5 | Z.AI | Open | 69.8% | |
| 7 | Qwen3.6-35B-A3B | Alibaba | Open | 60.1% | |