Most cost-efficient GPT-5.6 model for responsive, high-volume production workloads.
| Benchmark | Score | Rank |
|---|---|---|
MMMUvals.ai College-level multimodal reasoning across 30+ disciplines | 85% | #12 / 46 |
GPQAArtificial Analysis PhD-level science questions even experts struggle with | 91.1% | #16 / 73 |
hleArtificial Analysis | 37.2% | #16 / 61 |