| question-answering-on-drop-3-shot | DeepSeek-V3 | #2 | F1: 91.6 |
| code-generation-on-aider-polyglot | DeepSeek-V3 | #5 | Accuracy: 49.6 |
| instruction-following-on-ifeval | DeepSeek-V3 | #5 | Prompt-level strict-accuracy: 86.1 |
| question-answering-on-c-simpleqa | DeepSeek-V3 | #5 | Accuracy: 64.8 |
| question-answering-on-simpleqa | DeepSeek-V3 | #5 | Accuracy: 24.9 |
| question-answering-on-frames | DeepSeek-V3 | #6 | Accuracy: 73.3 |
| long-context-understanding-on-longbench-v2 | DeepSeek-V3 | #14 | Accuracy: 48.7 |
| multi-task-language-understanding-on-c-eval | DeepSeek-V3 | #18 | Accuracy: 86.5 |
| multi-task-language-understanding-on-mmlu-redux | DeepSeek-V3 | #19 | Accuracy: 89.1 |
| code-generation-on-humaneval | DeepSeek V3 | #38 | Pass@1: 87.2 |
| multi-task-language-understanding-on-mmlu-pro | DeepSeek-V3 | #54 | Accuracy: 75.9 |
| mathematical-reasoning-on-math-500 | DeepSeek-V3 | #59 | Accuracy: 90.2 |
| code-generation-on-swe-bench-verified | DeepSeek-V3 (Agentless) | #78 | Accuracy: 42.0 |
| mathematical-reasoning-on-aime24 | DeepSeek-V3 | #95 | Acc: 39.2 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | DeepSeek-V3 | #96 | Accuracy: 59.1 |