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: Failure Count ↑.
Failure Reasons
86/86
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) |
|---|---|---|---|---|---|---|---|
| #13 | GPT-5.3-Codex medium | OpenAI | 1 | 9.0 | $0.920 | 2/3 | 5.05s |
| #16 | Muse Spark 1.1 medium | Meta | 1 | 7.9 | $1.357 | 2/3 | 42.5s |
| #18 | GPT-5.4 medium | OpenAI | 1 | 8.2 | $1.533 | 2/3 | 9.14s |
| #21 | GPT-5.2 medium | OpenAI | 1 | 7.5 | $0.951 | 2/3 | 5.80s |
| #24 | Muse Spark 1.1 low | Meta | 1 | 8.3 | $0.647 | 2/3 | 6.60s |
| #26 | GPT-5 Mini medium | OpenAI | 1 | 5.6 | $0.237 | 1/3 | 15.2s |
| #28 | Inkling high | Thinkingmachines | 1 | 6.9 | $1.006 | 1/3 | 10.7s |
| #31 | GLM 5.2 high | Z.ai | 1 | 6.0 | $0.970 | 1/3 | 33.7s |
| #35 | Seed-2.0-Lite medium | Bytedance Seed | 1 | 9.0 | $0.234 | 2/3 | 10.2s |
| #45 | DeepSeek V4 Flash high | DeepSeek | 1 | 8.2 | $0.042 | 2/3 | 26.1s |
| #46 | DeepSeek V4 Pro high | DeepSeek | 1 | 6.9 | $0.200 | 1/3 | 56.8s |
| #49 | GLM 5 Turbo medium | Z.ai | 1 | 8.7 | $0.323 | 2/3 | 5.23s |
| #52 | Kimi K2.7 Code medium | Moonshot AI | 1 | 5.9 | $0.751 | 1/3 | 41.0s |
| #53 | GPT-5.4 Nano medium | OpenAI | 1 | 4.1 | $0.138 | 0/3 | 3.79s |
| #56 | GPT-5.4 Mini medium | OpenAI | 1 | 7.8 | $0.756 | 2/3 | 4.37s |