Instructions following: Wrong answer
Instructions following
Wrong answer
See which AI models are most likely to hit Wrong answer on Instructions following, so you can spot weak points faster. Sort by: Total Cost ↓.
Failure Reasons
111/111
Filter models
No models match the current search and filters.
| Rank | Model | Company | Wrong answer Count | Category Score | Total Cost | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|---|
| #269 | Inkling none | Thinkingmachines | 1 | 0.6 | $0.292 | 1/2 | 1.72s |
| #295 | Mistral Large 4 none | Mistral | 1 | 0.7 | $0.290 | 1/2 | 1.16s |
| #324 | Kimi K2.5 none | Moonshot AI | 1 | 0.7 | $0.262 | 1/2 | 2.67s |
| #270 | Qwen3.5 Plus 2026-04-20 none | Qwen | 1 | 0.6 | $0.216 | 1/2 | 1.17s |
| #385 | GLM 4.7 Flash medium | Z.ai | 1 | 0.6 | $0.193 | 1/2 | 2.97s |
| #128 | Qwen3.7 Plus none | Qwen | 1 | 0.6 | $0.178 | 1/2 | 929ms |
| #294 | GPT-5.4 Mini none | OpenAI | 1 | 0.6 | $0.170 | 1/2 | 728ms |
| #354 | Hy4 preview none | Tencent | 1 | 0.7 | $0.165 | 1/2 | 47.9s |
| #233 | Gemini 3 Flash Preview none | 1 | 0.6 | $0.155 | 1/2 | 1.58s | |
| #246 | Qwen3.6 27B none | Qwen | 1 | 0.6 | $0.148 | 1/2 | 1.92s |
| #349 | Laguna S 2.1 high | Poolside | 1 | 0.5 | $0.145 | 0/2 | 453ms |
| #283 | Qwen3.6 Flash none | Qwen | 1 | 0.6 | $0.139 | 1/2 | 1.10s |
| #338 | MiMo-V2.5-Pro none | Xiaomi | 1 | 0.6 | $0.135 | 1/2 | 1.03s |
| #334 | Mistral Small 4 medium | Mistral | 1 | 0.7 | $0.126 | 1/2 | 1.38s |
| #304 | Qwen3.5-35B-A3B none | Qwen | 1 | 0.6 | $0.120 | 1/2 | 809ms |