KORGym - Puzzle — leaderboard
Metric: Score. Source: razor233.github.io. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (03-25) | 0.94 |
| 2 | Claude 3.7 Sonnet (Thinking) | 0.93 |
| 3 | O1 (2024-12-17) | 0.83 |
| 4 | DeepSeek R1 | 0.82 |
| 5 | O3 Mini | 0.81 |
| 6 | DeepSeek V3 (0324) | 0.55 |
| 7 | Qwen 3 32B (Thinking) | 0.55 |
| 8 | Gemini 2.0 Flash (Thinking) | 0.53 |
| 9 | Gemini 2.0 Flash | 0.28 |
| 10 | DeepSeek R1 Distill Qwen 32B | 0.28 |
| 11 | GPT-4o | 0.25 |
| 12 | Doubao-1.5-Pro | 0.16 |
| 13 | Qwen 2.5 72B Instruct | 0.1 |
| 14 | Qwen 2.5 32B Instruct | 0.07 |
| 15 | DeepSeek-R1-Distill-Qwen-7B | 0.02 |
Interactive version: theaggregate.ai/benchmark?slug=korgym-puzzle · How the rankings work · Data refreshed daily, snapshot 2026-07-22.