CortexBench - Comparison (Closed-Book): leaderboard
Metric: Accuracy (%, GPT-5.4 judge). Source: arxiv.org. Saturation forecast: Around 2030. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 32.3 |
| 2 | Claude Opus 4.6 | 31.7 |
| 3 | Gemini 3 Pro | 28.6 |
| 4 | Claude Sonnet 4.6 | 25 |
| 5 | DeepSeek V4 Flash | 24.8 |
| 6 | Gemini 3.1 Pro (Preview) | 24.3 |
| 7 | DeepSeek V3.2 | 22.6 |
| 8 | GPT-4o | 21.6 |
Interactive version: theaggregate.ai/benchmark?slug=cortexbench-comparison-closed-book · How It Works · Data refreshed daily, snapshot 2026-09-25.