GPT-5.6 Luna (Low): benchmark results
GPT-5.6 Luna evaluated at the low reasoning-effort setting. Provider: OpenAI. Released 2026-07-09. Access: API.
Aggregate Rating 1667 ± 11, rank #466 of 2168 rated models, from 142 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| FinLifeBench - Financial State - Checkpoint State Accuracy | 77.1 | Cell-level state accuracy (%) | 90 |
| FinLifeBench - Financial State - Granular Change Accuracy | 46.1 | GCA@15 (%) | 90 |
| AA-Omniscience Index - Software Engineering (SWE) - Kotlin | 40 | Omniscience Index | 86.2 |
| AA-Omniscience Index - Software Engineering (SWE) - TypeScript | 47.78 | Omniscience Index | 83.9 |
| AA-Omniscience Index - Software Engineering (SWE) - PHP | 40 | Omniscience Index | 82.1 |
| AA-Omniscience Index - Software Engineering (SWE) - Go | 40 | Omniscience Index | 82 |
| AA-Omniscience Index - Software Engineering (SWE) - Python | 37.5 | Omniscience Index | 81.5 |
| AA-Omniscience Accuracy | 39.62 | Accuracy (%) | 78.2 |
| AA-Omniscience Index - Software Engineering (SWE) - JavaScript | 34.55 | Omniscience Index | 78.1 |
| AA-Omniscience Index - Software Engineering (SWE) | 28.4 | Omniscience Index | 77.7 |
| AA-Omniscience Index - Software Engineering (SWE) - C | 49 | Omniscience Index | 77.5 |
| Chess APO Bench | 31.72 | Puzzle accuracy (%) on the 559-puzzle Lichess test split wit | 75 |
Interactive version: theaggregate.ai/model?slug=gpt-5-6-luna-low · How It Works · Data refreshed daily, snapshot 2026-10-20.