| infographic-visual-question-answering-on-infovqa | Gemma 4 31B | #1 | ANLS: 92.0 |
| infographic-visual-question-answering-on-infovqa | Gemma 4 26B-A4B | #3 | ANLS: 89.3 |
| long-context-understanding-on-mrcr-v2-128k | Gemma 4 31B | #2 | Accuracy: 66.4 |
| long-context-understanding-on-mrcr-v2-128k | Gemma 4 12B Unified | #4 | Accuracy: 43.4 |
| multi-task-language-understanding-on-mmmlu | Gemma 4 31B | #2 | Accuracy: 88.4 |
| multi-task-language-understanding-on-mmmlu | Gemma 4 12B Unified | #9 | Accuracy: 83.4 |
| multimodal-reasoning-on-mathvision | Gemma 4 31B | #6 | Accuracy: 85.6 |
| multimodal-reasoning-on-mathvision | Gemma 4 26B-A4B | #9 | Accuracy: 82.4 |
| multimodal-reasoning-on-mathvision | Gemma 4 12B Unified | #10 | Accuracy: 79.7 |
| multimodal-reasoning-on-mmmu-pro | Gemma 4 31B | #8 | Accuracy: 76.9 |
| multimodal-reasoning-on-mmmu-pro | Gemma 4 26B-A4B | #15 | Accuracy: 73.8 |
| multimodal-reasoning-on-mmmu-pro | Gemma 4 12B Unified | #21 | Accuracy: 69.1 |
| agents-on-bench | Gemma 4 31B | #15 | Accuracy: 76.9 |
| agents-on-bench | Gemma 4 12B Unified | #21 | Accuracy: 69.0 |
| multi-task-language-understanding-on-mmlu-pro | Gemma 4 31B | #15 | Accuracy: 85.2 |
| multi-task-language-understanding-on-mmlu-pro | Gemma 4 26B-A4B | #31 | Accuracy: 82.6 |
| multi-task-language-understanding-on-mmlu-pro | Gemma 4 12B Unified | #52 | Accuracy: 77.2 |
| multi-task-language-understanding-on-mmlu-pro | Gemma 4 E4B | #75 | Accuracy: 69.4 |
| multi-task-language-understanding-on-mmlu-pro | Gemma 4 E2B | #96 | Accuracy: 60.0 |
| code-generation-on-livecodebench-v6 | Gemma 4 31B | #20 | Pass@1: 80.0 |
| code-generation-on-livecodebench-v6 | Gemma 4 12B Unified | #31 | Pass@1: 72.0 |
| mathematical-reasoning-on-aime-2026 | Gemma 4 31B | #21 | Accuracy: 89.2 |
| mathematical-reasoning-on-aime-2026 | Gemma 4 12B Unified | #31 | Accuracy: 77.5 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Gemma 4 31B | #40 | Accuracy: 84.3 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Gemma 4 26B-A4B | #49 | Accuracy: 82.3 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Gemma 4 12B Unified | #58 | Accuracy: 78.8 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Gemma 4 E4B | #98 | Accuracy: 58.6 |
| question-answering-on-humanity-s-last-exam-hle | Gemma 4 26B-A4B | #41 | Accuracy: 8.7 |
| question-answering-on-humanity-s-last-exam-hle | Gemma 4 12B Unified | #42 | Accuracy: 5.2 |