SuperGrok Heavy beta with native video understanding. AA now tracking with full per-benchmark coverage; #2 globally on initial AA pass.
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 |
|---|---|---|
Arena EloArtificial Analysis Human preference ranking via blind comparisons | 1498 | #11 / 51 |
MMMUvals.ai College-level multimodal reasoning across 30+ disciplines | 83.1% | #30 / 51 |
LiveCodeBenchvals.ai Contamination-free competitive programming (filtered by cutoff date) | 84.5% | #31 / 62 |
TerminalArtificial Analysis Agentic terminal coding tasks requiring multi-step execution | 37.9% | #36 / 58 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 90.1% | #40 / 96 |
MMLU-Provals.ai Harder 10-option successor to MMLU; more reasoning-focused | 85.8% | #43 / 61 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 37.2% | #51 / 100 |
| Benchmark / source | Score |
|---|---|
HLE Artificial AnalysisGrok 4.3 (High) | 37.2% |
MMMU-Pro Artificial AnalysisGrok 4.3 (High) | 78.1% |
Terminal-Bench 2.1 Artificial AnalysisGrok 4.3 (High) | 39.7% |
Terminal-Bench 4.0 Artificial AnalysisGrok 4.3 (High) | 0% |
SciCode Artificial AnalysisGrok 4.3 (High) | 48.3% |
AA-LCR Artificial AnalysisGrok 4.3 (High) | 73% |
CritPt Artificial AnalysisGrok 4.3 (High) | 8% |
ITBench SRE Artificial AnalysisGrok 4.3 (High) | 32.7% |
IFBench Artificial AnalysisGrok 4.3 (High) | 81.3% |