Google's first natively multimodal model, launched the Gemini brand
| Benchmark | Score | Rank |
|---|---|---|
HellaSwag Common sense reasoning about everyday situations | 92.1% | #22 / 36 |
ARC-C Grade-school science questions requiring reasoning | 93.8% | #33 / 40 |
HumanEval Coding ability - generating correct Python functions | 74.3% | #41 / 49 |
MATH Competition-level mathematics problems | 58.6% | #41 / 49 |
MMLU Tests knowledge across 57 subjects from STEM to humanities | 83.7% | #45 / 53 |
GPQA PhD-level science questions even experts struggle with | 42.1% | #51 / 54 |