Reasoning at 75% lower cost than o1, made chain-of-thought economically viable
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 20 models released within six months of it, across 5 benchmarks where scores spread out.
| Benchmark | Score | Rank |
|---|---|---|
MATH Competition-level mathematics problems | 97.3% | #9 / 50 |
HumanEval Coding ability - generating correct Python functions | 93.1% | #23 / 50 |
Arena Elo Human preference ranking via blind comparisons | 1361 | #33 / 51 |
SWE-bench Real-world GitHub issue resolution | 49.3% | #34 / 39 |
MMLU Tests knowledge across 57 subjects from STEM to humanities | 87.5% | #35 / 54 |
TerminalArtificial Analysis Agentic terminal coding tasks requiring multi-step execution | 6.1% | #55 / 58 |
GPQA PhD-level science questions even experts struggle with | 77% | #74 / 96 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 12% | #87 / 100 |
| Benchmark / source | Score |
|---|---|
HLE Artificial Analysiso3-mini (High) | 12% |
Terminal-Bench 2.1 Artificial Analysiso3-mini (High) | 4.5% |
Terminal-Bench 4.0 Artificial Analysiso3-mini (High) | 0% |
SciCode Artificial Analysiso3-mini (High) | 42.8% |
AA-LCR Artificial Analysiso3-mini (High) | 43% |
CritPt Artificial Analysiso3-mini (High) | 0.3% |
IFBench Artificial Analysiso3-mini (High) | 67.1% |