European frontier model with strong multilingual and code performance
SI v3 uses weighted percentages from one Artificial Analysis configuration. Ranking requires all five benchmarks. Partial scores average only available weights. Scores are not comparable to SI v2 or a probability of reaching singularity.
Compared with 13 models released within six months of it, across 3 benchmarks where scores spread out.
| Benchmark | Score | Rank |
|---|---|---|
HellaSwag Common sense reasoning about everyday situations | 89.4% | #24 / 36 |
HumanEval Coding ability - generating correct Python functions | 92.1% | #29 / 50 |
ARC-C Grade-school science questions requiring reasoning | 94.2% | #32 / 40 |
MATH Competition-level mathematics problems | 69.2% | #39 / 50 |
MMLU Tests knowledge across 57 subjects from STEM to humanities | 84% | #45 / 54 |
Arena Elo Human preference ranking via blind comparisons | 1178 | #50 / 51 |
TerminalArtificial Analysis Agentic terminal coding tasks requiring multi-step execution | 6.1% | #57 / 58 |
GPQA PhD-level science questions even experts struggle with | 46.3% | #91 / 96 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 3.3% | #98 / 100 |
| Benchmark / source | Score |
|---|---|
HLE Artificial AnalysisMistral Large 2 (Nov '24) | 3.3% |
CritPt Artificial AnalysisMistral Large 2 (Nov '24) | 0% |
IFBench Artificial AnalysisMistral Large 2 (Nov '24) | 31.2% |