Reasoning-focused flagship, competitive with frontier Western models
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.
| Benchmark | Score | Rank |
|---|---|---|
MMLU-Provals.ai Harder 10-option successor to MMLU; more reasoning-focused | 87% | #33 / 61 |
LiveCodeBenchvals.ai Contamination-free competitive programming (filtered by cutoff date) | 81.8% | #40 / 62 |
TerminalArtificial Analysis Agentic terminal coding tasks requiring multi-step execution | 28.8% | #44 / 58 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 83% | #68 / 96 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 23.2% | #76 / 100 |
| Benchmark / source | Score |
|---|---|
HLE Artificial AnalysisMiniMax-M2.1 | 23.2% |
AA-LCR Artificial AnalysisMiniMax-M2.1 | 67.7% |
CritPt Artificial AnalysisMiniMax-M2.1 | 0.3% |
IFBench Artificial AnalysisMiniMax-M2.1 | 69.9% |