GPT-6 Luna (xHigh): benchmark results
Provider: OpenAI. Released 2026-09-22. Access: API.
Aggregate Rating 1781 ± 11, rank #185 of 2178 rated models, from 133 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MLCR-AA - Accuracy | 86.67 | Accuracy (%) | 89.1 |
| AA CritPt | 17.43 | Accuracy (%) | 85.6 |
| AA Long Context Reasoning | 80 | Accuracy (%) | 85 |
| AA Humanity's Last Exam | 34.34 | Accuracy (%) | 79.8 |
| AA GDP.pdf - HR - Criterion Pass Rate | 77.27 | Criterion Pass Rate (%) | 79.2 |
| AA MMMU-Pro | 79.71 | Accuracy (%) | 77.9 |
| ObviousBench | 97.92 | Answer pass³ (%) | 77 |
| o11y-bench - Pass^3 | 84.38 | Tasks passed on all three attempts, Pass^3 (%) | 75.7 |
| AA GDP.pdf - Healthcare - Criterion Pass Rate | 82.97 | Criterion Pass Rate (%) | 73.9 |
| MLCR-AA | 20.56 | Overall Score (%) | 73.4 |
| AA GDP.pdf - Manufacturing/Supply Chains - Criterion Pass Rate | 70.8 | Criterion Pass Rate (%) | 73.1 |
| AA GDP.pdf - Engineering - Criterion Pass Rate | 67.62 | Criterion Pass Rate (%) | 72.2 |
Interactive version: theaggregate.ai/model?slug=gpt-6-luna-xhigh · How It Works · Data refreshed daily, snapshot 2026-11-07.