First Claude with 100K context, established Anthropic as a frontier lab
| Benchmark | Score | Rank |
|---|---|---|
HellaSwag Common sense reasoning about everyday situations | 89.1% | #26 / 36 |
ARC-C Grade-school science questions requiring reasoning | 93.2% | #34 / 40 |
HumanEval Coding ability - generating correct Python functions | 71.2% | #42 / 49 |
MATH Competition-level mathematics problems | 42.6% | #46 / 49 |
MMLU Tests knowledge across 57 subjects from STEM to humanities | 78.5% | #48 / 53 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 34.4% | #54 / 54 |