ProcArena - Oracle - Direct - Code Composition: leaderboard
Metric: Execution Accuracy (%). Source: arxiv.org. Saturation forecast: Around 2031. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 36.4 |
| 2 | GPT-5.5 | 36.4 |
| 3 | GPT-5.2 | 34.7 |
| 4 | GPT-5.4 Mini | 33.9 |
| 5 | DeepSeek V4 Pro | 30.5 |
| 6 | GLM-5.2 | 28 |
| 7 | GLM-5.3 | 28 |
Interactive version: theaggregate.ai/benchmark?slug=procarena-oracle-direct-code-composition · How It Works · Data refreshed daily, snapshot 2026-09-24.