Puzzle Solving: Did not follow instructions
Puzzle Solving
Did not follow instructions
See which AI models are most likely to hit Did not follow instructions on Puzzle Solving, so you can spot weak points faster. Sort by: Total Cost ↑.
Failure Reasons
157/157
Filter models
No models match the current search and filters.
| Rank | Model | Company | Did not follow instructions Count | Category Score | Total Cost | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|---|
| #168 | Qwen3.5-27B none | Qwen | 1 | 0.7 | $0.116 | 1/3 | 1.38s |
| #313 | Qwen3.5-35B-A3B none | Qwen | 1 | 0.4 | $0.120 | 0/3 | 1.35s |
| #223 | Mercury 2 medium | Inception | 1 | 0.5 | $0.121 | 1/3 | 949ms |
| #343 | Mistral Small 4 medium | Mistral | 1 | 0.3 | $0.126 | 0/3 | 2.17s |
| #347 | MiMo-V2.5-Pro none | Xiaomi | 1 | 0.7 | $0.135 | 1/3 | 1.30s |
| #291 | Qwen3.6 Flash none | Qwen | 1 | 0.4 | $0.139 | 0/3 | 1.21s |
| #39 | Claude Haiku 5.5 xhigh | Anthropic | 1 | 0.8 | $0.140 | 2/3 | 3.80s |
| #253 | Qwen3.6 27B none | Qwen | 1 | 0.5 | $0.148 | 1/3 | 5.15s |
| #106 | DeepSeek V4 Pro 0423 high | DeepSeek | 1 | 0.7 | $0.149 | 1/3 | 56.8s |
| #265 | DeepSeek V4 Flash 0423 none | DeepSeek | 1 | 0.3 | $0.149 | 0/3 | 23.7s |
| #263 | DeepSeek V3.2 medium | DeepSeek | 1 | 0.7 | $0.152 | 1/3 | 37.7s |
| #302 | GPT-5 Nano medium | OpenAI | 1 | 0.5 | $0.159 | 1/3 | 20.6s |
| #293 | Nemotron 3 Ultra none | NVIDIA | 1 | 0.6 | $0.164 | 1/3 | 1.06s |
| #363 | Hy4 preview none | Tencent | 1 | 0.5 | $0.165 | 0/3 | 24.1s |
| #339 | Nemotron 3.5 Lightning high | NVIDIA | 2 | 0.4 | $0.168 | 0/3 | 6.09s |