Anthropic's strongest generally available model for long-running coding, research, and knowledge work, with Mythos-level capability and adaptive reasoning.
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 60 models released within six months of it, across 6 benchmarks where scores spread out.
| Benchmark | Score | Rank |
|---|---|---|
MMLU-Provals.ai Harder 10-option successor to MMLU; more reasoning-focused | 92.4% | #1 / 61 |
LiveCodeBenchvals.ai Contamination-free competitive programming (filtered by cutoff date) | 90.5% | #1 / 62 |
MMMUvals.ai College-level multimodal reasoning across 30+ disciplines | 90.6% | #1 / 51 |
ARC-AGIARC Prize Novel reasoning tasks requiring fluid intelligence | 99.2% | #1 / 39 |
HLEArtificial Analysis Challenging multidisciplinary questions evaluated by Artificial Analysis | 59.1% | #2 / 95 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 93.7% | #8 / 95 |
| Benchmark / source | Score |
|---|---|
HLE Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 59.1% |
Terminal-Bench 2.1 Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 91.4% |
Terminal-Bench 4.0 Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 52% |
SciCode Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 63.1% |
AA-LCR Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 85.3% |
CritPt Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 29.7% |
ITBench SRE Artificial AnalysisClaude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 49.5% |