Small open-weight Qwen 3.8 tier, holding GPQA 90.5 at 27B parameters.
| Benchmark | Score | Rank |
|---|---|---|
aaAgentic | 50.9% | #6 / 12 |
AA Intelligence Independent composite of agentic work, coding, scientific reasoning, and general capability | 52% | #14 / 18 |
AA Coding Independent composite of implementation, terminal use, and real-codebase agent performance | 68.1% | #17 / 18 |
MMMUvals.ai College-level multimodal reasoning across 30+ disciplines | 83.9% | #22 / 57 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 90.5% | #29 / 88 |
LiveCodeBenchvals.ai Contamination-free competitive programming (filtered by cutoff date) | 84% | #31 / 59 |
hleArtificial Analysis | 33.9% | #35 / 76 |
MMLU-Provals.ai Harder 10-option successor to MMLU; more reasoning-focused | 84.3% | #43 / 58 |