CocoaBench — leaderboard
Metric: Accuracy. Source: cocoabench.github.io. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 (High) | 45.1 |
| 2 | Claude Sonnet 4.6 (High) | 34 |
| 3 | Gemini 3.1 Pro (Preview) | 30.7 |
| 4 | Kimi K2.5 | 11.8 |
Interactive version: theaggregate.ai/benchmark?slug=cocoabench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.