Puzzle Solving Ranking
See which AI models perform best on Puzzle Solving, which ones stay reliable, and where the biggest gaps appear. Sort by: Total Cost ↓.
451/451
Filter models
No models match the current search and filters.
| Rank | Model | Company | Puzzle Solving Score | Score | Total Cost | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|---|
| #195 | Gemini 3.1 Flash Lite medium | 0.8 | 0.7 | $0.153 | 2/3 | 1.95s | |
| #263 | DeepSeek V3.2 medium | DeepSeek | 0.7 | 0.6 | $0.152 | 1/3 | 37.7s |
| #265 | DeepSeek V4 Flash 0423 none | DeepSeek | 0.3 | 0.6 | $0.149 | 0/3 | 23.7s |
| #106 | DeepSeek V4 Pro 0423 high | DeepSeek | 0.7 | 0.8 | $0.149 | 1/3 | 56.8s |
| #253 | Qwen3.6 27B none | Qwen | 0.5 | 0.6 | $0.148 | 1/3 | 5.15s |
| #358 | Laguna S 2.1 high | Poolside | 0.3 | 0.5 | $0.145 | 0/3 | 1.62s |
| #197 | Gemma 4 31B medium | 1.0 | 0.7 | $0.144 | 3/3 | 26.9s | |
| #148 | Seed-2.0-Mini medium | Bytedance Seed | 0.8 | 0.8 | $0.140 | 2/3 | 31.8s |
| #312 | GLM 5 none | Z.ai | 0.8 | 0.6 | $0.140 | 2/3 | 1.91s |
| #39 | Claude Haiku 5.5 xhigh | Anthropic | 0.8 | 0.9 | $0.140 | 2/3 | 3.80s |
| #291 | Qwen3.6 Flash none | Qwen | 0.4 | 0.6 | $0.139 | 0/3 | 1.21s |
| #365 | Hy3 preview high | Tencent | 0.8 | 0.5 | $0.135 | 2/3 | 27.9s |
| #347 | MiMo-V2.5-Pro none | Xiaomi | 0.7 | 0.5 | $0.135 | 1/3 | 1.30s |
| #183 | MiMo-V2.5 medium | Xiaomi | 0.8 | 0.7 | $0.130 | 2/3 | 20.3s |
| #343 | Mistral Small 4 medium | Mistral | 0.3 | 0.5 | $0.126 | 0/3 | 2.17s |