BinaryAudit: leaderboard
Metric: Avg Success Rate (%). Source: quesma.com. 26 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.3 Codex | 94.12 |
| 2 | GPT-5.2 Codex (xHigh) | 90.67 |
| 3 | GPT-5 | 88.89 |
| 4 | Grok 4.1 Fast | 86.11 |
| 5 | Gemini 3.1 Pro (Preview) | 84.62 |
| 6 | GPT-5.4 | 84.44 |
| 7 | DeepSeek V3.2 | 83.33 |
| 8 | GPT-5.3 Codex (High) | 82.46 |
| 9 | Claude Opus 4.6 | 82.05 |
| 10 | GPT-5.4 (High) | 81.67 |
| 11 | GPT-5.3 Codex (xHigh) | 79.49 |
| 12 | Claude Sonnet 4.5 | 79.37 |
| 13 | Claude Opus 4.5 | 76 |
| 14 | Kimi K2.5 | 75.93 |
| 15 | GPT-5.2 | 74.6 |
Interactive version: theaggregate.ai/benchmark?slug=binaryaudit · How It Works · Data refreshed daily, snapshot 2026-09-05.