Puzzle Solving Ranking
See which AI models perform best on Puzzle Solving, which ones stay reliable, and where the biggest gaps appear. Sort by: Metric ↑.
311/311
Filter models
No models match the current search and filters.
| Rank | Model | Company | Puzzle Solving Score | Score | Total Cost | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|---|
| #72 | Gemini 3.5 Flash Lite medium | 8.2 | 7.8 | $0.272 | 2/3 | 1.82s | |
| #73 | GLM 5.2 medium | Z.ai | 8.2 | 7.8 | $0.224 | 2/3 | 13.1s |
| #84 | DeepSeek V4 Flash 0423 high | DeepSeek | 8.2 | 7.6 | $0.039 | 2/3 | 26.1s |
| #109 | GLM 5.3 high | Z.ai | 8.2 | 7.3 | $0.403 | 2/3 | 4.74s |
| #117 | KAT-Coder-Pro V2.5 high | Kwaipilot | 8.2 | 7.2 | $0.506 | 2/3 | 3.38s |
| #119 | Qwen3.5 Plus 2026-04-20 medium | Qwen | 8.2 | 7.2 | $0.323 | 2/3 | 17.7s |
| #129 | GLM 5.1 medium | Z.ai | 8.2 | 7.0 | $0.727 | 2/3 | 31.6s |
| #133 | Seed-2.0-Mini medium | Bytedance Seed | 8.2 | 7.0 | $0.116 | 2/3 | 31.8s |
| #140 | Qwen3.6 Flash medium | Qwen | 8.2 | 6.8 | $0.741 | 2/3 | 6.29s |
| #142 | Seed-2.0-Code medium | Bytedance Seed | 8.2 | 6.8 | $1.153 | 2/3 | 34.3s |
| #151 | KAT-Coder-Pro V2.5 none | Kwaipilot | 8.2 | 6.7 | $0.487 | 2/3 | 3.03s |
| #173 | MiMo-V2.5 medium | Xiaomi | 8.2 | 6.4 | $0.104 | 2/3 | 20.3s |
| #183 | Qwen3.5-35B-A3B medium | Qwen | 8.2 | 6.1 | $1.140 | 2/3 | 33.1s |
| #187 | Qwen3.5-Flash medium | Qwen | 8.2 | 6.1 | $0.142 | 2/3 | 27.6s |
| #260 | Mercury 2.5 Preview low | Inception | 8.2 | 5.0 | $0.011 | 2/3 | 710ms |