Open-weight multimodal MoE optimized for efficient local and hosted deployment.
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 models released within six months of it.
| Benchmark | Score | Rank |
|---|---|---|
ARC-AGIARC Prize Novel reasoning tasks requiring fluid intelligence | 40.1% | #28 / 40 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 89.5% | #44 / 96 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 33.3% | #64 / 102 |
| Benchmark / source | Score |
|---|---|
HLE Artificial AnalysisInkling Small | 33.3% |
MMMU-Pro Artificial AnalysisInkling Small | 74% |
Terminal-Bench 2.1 Artificial AnalysisInkling Small | 55.1% |
Terminal-Bench 4.0 Artificial AnalysisInkling Small | 1% |
SciCode Artificial AnalysisInkling Small | 49.7% |
AA-LCR Artificial AnalysisInkling Small | 75.7% |
CritPt Artificial AnalysisInkling Small | 8.3% |