Frontier Anthropic model with strong coding, reasoning, and professional-work 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 79 models released within six months of it, across 7 benchmarks where scores spread out.
| Benchmark | Score | Rank |
|---|---|---|
Terminal Agentic terminal coding tasks requiring multi-step execution | 78.9% | #7 / 58 |
OSWorld Computer use in real desktop environments | 54.8% | #9 / 13 |
MMLU-Provals.ai Harder 10-option successor to MMLU; more reasoning-focused | 89.6% | #9 / 61 |
LiveCodeBenchvals.ai Contamination-free competitive programming (filtered by cutoff date) | 87.8% | #11 / 62 |
AA Coding (Historical) Archived release-era coding composite, excluded from current rankings because comparable current-series coverage is unavailable | 72.5% | #13 / 22 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 48.7% | #13 / 100 |
AA Intelligence (Historical) Archived release-era composite, not comparable to the current versioned Intelligence Index used for rankings | 55.7% | #14 / 22 |
ARC-AGIARC Prize Novel reasoning tasks requiring fluid intelligence | 72.1% | #15 / 40 |
MMMUvals.ai College-level multimodal reasoning across 30+ disciplines | 86.6% | #16 / 51 |
GPQA PhD-level science questions even experts struggle with | 92% | #27 / 96 |
| Benchmark / source | Score |
|---|---|
HLE Artificial AnalysisClaude Opus 4.8 (Max) | 48.7% |
Terminal-Bench 2.1 Artificial AnalysisClaude Opus 4.8 (Max) | 84.6% |
Terminal-Bench 4.0 Artificial AnalysisClaude Opus 4.8 (Max) | 21.7% |
SciCode Artificial AnalysisClaude Opus 4.8 (Max) | 54.4% |
AA-LCR Artificial AnalysisClaude Opus 4.8 (Max) | 77.7% |
CritPt Artificial AnalysisClaude Opus 4.8 (Max) | 20.9% |
IFBench Artificial AnalysisClaude Opus 4.8 (Max) | 62.2% |