| Rank | Model | Score |
|---|---|---|
| 1 | claude-fable-5 | 93.3 |
| 2 | gemini-3-1-pro | 92.7 |
| 3 | gpt-5-6-sol | 91.8 |
| 4 | deepseek-v4-pro | 89.3 |
| 5 | glm-5-2 | 89.2 |
| 6 | inkling | 88.7 |
| 7 | gemini-2-5-pro | 88.6 |
| 8 | kimi-k2-6 | 88.4 |
| 9 | nemotron-3-ultra-550b-a55b | 85.6 |
| 10 | kimi-k2-5 | 84 |
| 11 | gemini-2-0-flash-lite | 78.2 |
| 12 | gemma-3-27b | 75.1 |
| 13 | gemma-3-12b | 69.5 |
| 14 | gemma-3n-e4b-litert-preview | 64.5 |
| 15 | gemma-3n-e4b | 64.5 |
| 16 | gemma-3n-e2b-litert-preview | 59 |
| 17 | gemma-3n-e2b | 59 |
| 18 | gemma-3-4b | 54.5 |
| 19 | gemma-3-1b | 34.2 |
1 phaseActive