AI BENCHY Failures
Did not follow instructions Failures
See which AI models run into Did not follow instructions most often, so you can spot reliability risks before choosing one. Sort by: Failure Count ↑.
| Rank | Model | Company | Did not follow instructions Count | Score | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|
| #42 | Claude Sonnet 4.6 none | Anthropic | 1 | 7.4 | 11/18 | 4.98s |
| #48 | Gemma 4 31B none | 1 | 6.9 | 10/18 | 4.02s | |
| #59 | Qwen3.5-Flash none | Qwen | 1 | 6.2 | 8/18 | 3.25s |
| #62 | Gemini 2.5 Flash none | 1 | 6.2 | 7/18 | 903ms | |
| #66 | GPT-5.4 none | OpenAI | 1 | 5.9 | 7/18 | 1.51s |
| #70 | Qwen3.5-122B-A10B none | Qwen | 1 | 5.7 | 6/18 | 3.69s |
| #87 | Qwen3 Coder Next none | Qwen | 1 | 5.1 | 4/18 | 10.2s |
| #89 | GPT-4o-mini none | OpenAI | 1 | 4.9 | 4/18 | 2.00s |
| #91 | Mercury 2 none | Inception | 1 | 4.8 | 4/18 | 613ms |
| #94 | MiMo-V2-Flash none | Xiaomi | 1 | 4.5 | 3/18 | 2.79s |
| #6 | Seed-2.0-Lite medium | Bytedance Seed | 2 | 8.6 | 13/18 | 30.4s |
| #7 | GPT-5.3-Codex medium | OpenAI | 2 | 8.6 | 13/18 | 15.4s |
| #10 | Qwen3.5-27B medium | Qwen | 2 | 8.4 | 13/18 | 53.0s |
| #16 | GPT-5.4 medium | OpenAI | 2 | 8.2 | 13/18 | 18.6s |
| #18 | GLM 5 Turbo medium | Z.ai | 2 | 8.1 | 12/18 | 17.7s |