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
121/121
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) |
|---|---|---|---|---|---|---|---|
| #149 | Solar Pro 4 medium | Upstage | 1 | 9.0 | $0.047 | 2/3 | 45.8s |
| #187 | Gemini 3.1 Flash Lite minimal | 2 | 6.0 | $0.047 | 1/3 | 2.15s | |
| #262 | GLM 5 Turbo none | Z.ai | 1 | 5.5 | $0.047 | 1/3 | 2.65s |
| #130 | Solar Pro 4 high | Upstage | 1 | 8.7 | $0.046 | 2/3 | 55.2s |
| #191 | Gemini 3.1 Flash Lite none | 1 | 6.3 | $0.046 | 1/3 | 720ms | |
| #240 | KAT-Coder-Air V2.5 low | Kwaipilot | 1 | 3.1 | $0.046 | 0/3 | 1.57s |
| #229 | Mimo V2 PRO none | Xiaomi | 1 | 6.0 | $0.045 | 1/3 | 1.61s |
| #66 | DeepSeek V4 Flash 0731 low | DeepSeek | 1 | 8.0 | $0.044 | 2/3 | 9.08s |
| #148 | Solar Pro 4 low | Upstage | 1 | 9.0 | $0.044 | 2/3 | 42.1s |
| #125 | Qwen3.7 Flash low | Qwen | 1 | 7.3 | $0.043 | 1/3 | 9.79s |
| #89 | DeepSeek V4 Flash 0423 high | DeepSeek | 1 | 8.2 | $0.042 | 2/3 | 26.1s |
| #276 | GPT-5.4 Nano none | OpenAI | 1 | 5.4 | $0.041 | 1/3 | 1.25s |
| #273 | Hy4 preview none | Tencent | 1 | 4.6 | $0.041 | 0/3 | 24.1s |
| #238 | DeepSeek V4 Flash 0423 none | DeepSeek | 1 | 3.1 | $0.039 | 0/3 | 23.7s |
| #288 | Qwen3 Coder Next medium | Qwen | 1 | 3.0 | $0.034 | 0/3 | 1.25s |