GPT-5.6 Luna (Medium): benchmark results
GPT-5.6 Luna evaluated at the medium reasoning-effort setting. Provider: OpenAI. Released 2026-07-09. Access: API.
Aggregate Rating 1716 ± 10, rank #311 of 2168 rated models, from 160 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HoF-Bench | 65 | CVEs rediscovered at pass@4 (out of 95), with the generated- | 100 |
| HoF-Bench (No Context) | 65 | CVEs rediscovered at pass@4 (out of 95), without the generat | 100 |
| MultiView-Bench - 3D DoF 1 | 63.7 | Accuracy (%; six views, objects share the central object axi | 100 |
| MultiView-Bench - 3D Real World | 60 | Accuracy (%; six views of 3DCoMPaT++ real-world object asset | 96.4 |
| MultiView-Bench - 3D DoF 2 | 44.3 | Accuracy (%; six views, objects share the central object pla | 92.9 |
| PLCC - History | 94 | Accuracy (%) | 89.8 |
| PLCC - Culture & Tradition | 93 | Accuracy (%) | 89.6 |
| PLCC - Overall | 89.67 | Mean category accuracy (%) | 88 |
| PLCC - Vocabulary | 87 | Accuracy (%) | 86.9 |
| AA-Omniscience Index - Software Engineering (SWE) - Swift | 48 | Omniscience Index | 84.3 |
| PLCC - Art & Entertainment | 83 | Accuracy (%) | 84.2 |
| PLCC - Grammar | 87 | Accuracy (%) | 84 |
Interactive version: theaggregate.ai/model?slug=gpt-5-6-luna-medium · How It Works · Data refreshed daily, snapshot 2026-10-20.