GPT-5.6 Luna (xHigh) — benchmark results
Provider: OpenAI. Released 2026-07-09. Access: API.
Unified ELO 1932 ± 42, rank #45 of 1744 rated models, from 34 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Omniscience - Software Engineering (SWE) - Kotlin | 70 | Accuracy (%) | 96.4 |
| Artificial Analysis Intelligence Index | 49.07 | Intelligence Index | 96.4 |
| AA CritPt | 20.57 | Accuracy (%) | 96.3 |
| AA Omniscience - Software Engineering (SWE) - TypeScript | 75.56 | Accuracy (%) | 95.2 |
| AA Humanity's Last Exam | 35.59 | Accuracy (%) | 94.2 |
| AA GPQA Diamond | 89.49 | Accuracy (%) | 93.1 |
| AA Omniscience - Software Engineering (SWE) - Go | 66 | Accuracy (%) | 93.1 |
| AA Omniscience - Software Engineering (SWE) - Python | 71 | Accuracy (%) | 93.1 |
| AA Long Context Reasoning | 69.67 | Accuracy (%) | 92.9 |
| AA SciCode | 50 | Accuracy (%) | 92.9 |
| AA Omniscience - Software Engineering (SWE) - PHP | 70 | Accuracy (%) | 92.7 |
| AA Omniscience - Software Engineering (SWE) - HTML | 72 | Accuracy (%) | 92.6 |
Interactive version: aibenchmarks.dev/model?slug=gpt-5-6-luna-xhigh · How the rankings work · Data refreshed daily, snapshot 2026-07-10.