First 1M token context window, processed entire codebases in one pass
| Benchmark | Score | Rank |
|---|---|---|
HellaSwag Common sense reasoning about everyday situations | 92.5% | #20 / 36 |
ARC-C Grade-school science questions requiring reasoning | 94.4% | #31 / 40 |
MMMUArtificial Analysis College-level multimodal reasoning across 30+ disciplines | 55% | #31 / 33 |
Arena Elo Human preference ranking via blind comparisons | 1260 | #35 / 41 |
HumanEval Coding ability - generating correct Python functions | 84.1% | #39 / 49 |
MATH Competition-level mathematics problems | 67.7% | #39 / 49 |
MMLU Tests knowledge across 57 subjects from STEM to humanities | 85.9% | #42 / 53 |
GPQA PhD-level science questions even experts struggle with | 46.2% | #50 / 54 |