Reasoning-focused flagship, competitive with frontier Western models
| Benchmark | Score | Rank |
|---|---|---|
MMLU-Provals.ai Harder 10-option successor to MMLU; more reasoning-focused | 87% | #31 / 58 |
LiveCodeBenchvals.ai Contamination-free competitive programming (filtered by cutoff date) | 81.8% | #38 / 59 |
TerminalArtificial Analysis Agentic terminal coding tasks requiring multi-step execution | 28.8% | #44 / 56 |
hleArtificial Analysis | 22.2% | #57 / 79 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 83% | #63 / 91 |