Puzzle Solving on PUZZLE Avg across difficulty levels
54.3AccuracyGPT-5
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-52026.06 | 54.3 | 12,867.2 | |
| DeepSeek V3.22026.06 | 43.8 | 23,038.1 | |
| Kimi K22026.06 | 43.8 | 36,343.4 | |
| Qwen 3 235BParameters=235B2026.06 | 33.8 | 19,000.7 |
| Method | Links | ||
|---|---|---|---|
| GPT-52026.06 | 54.3 | 12,867.2 | |
| DeepSeek V3.22026.06 | 43.8 | 23,038.1 | |
| Kimi K22026.06 | 43.8 | 36,343.4 | |
| Qwen 3 235BParameters=235B2026.06 | 33.8 | 19,000.7 |