OpenAI GPT-6 Sol & Luna Launch - Coding Deception: leaderboard
Metric: Deception rate (%). Source: openai.com. Saturation forecast: Estimated already saturated. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6 (Max) | 0.51 |
| 2 | GPT-6 Sol (Max) | 1.3 |
| 3 | GPT-6 Luna (Max) | 2.81 |
| 4 | GPT-5.6 Luna (Max) | 9.54 |
| 5 | GPT-5.6 Sol (Max) | 10.41 |
Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-6-sol-luna-launch-coding-deception · How It Works · Data refreshed daily, snapshot 2026-09-24.