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) |
|---|---|---|---|---|---|---|
| #147 | GPT-4o-mini none | OpenAI | 1 | 4.8 | 5/21 | 1.77s |
| #149 | Nemotron 3 Nano Omni 30b A3b Reasoning medium | NVIDIA | 1 | 4.6 | 4/19 | 17.1s |
| #155 | Mercury 2 none | Inception | 1 | 4.5 | 4/21 | 653ms |
| #160 | LFM2-24B-A2B none | Liquid | 1 | 4.2 | 2/16 | 782ms |
| #161 | Qwen3.5-9B medium | Qwen | 1 | 4.2 | 3/21 | 82.2s |
| #15 | GPT-5.3-Codex medium | OpenAI | 2 | 8.4 | 15/21 | 16.2s |
| #19 | Seed-2.0-Lite medium | Bytedance Seed | 2 | 8.2 | 14/21 | 47.1s |
| #21 | GPT-5.4 medium | OpenAI | 2 | 8.0 | 14/21 | 22.3s |
| #30 | Qwen3.5-27B medium | Qwen | 2 | 7.8 | 13/21 | 68.4s |
| #31 | DeepSeek V4 Flash high | DeepSeek | 2 | 7.7 | 13/21 | 45.8s |
| #38 | Grok 4.3 medium | X AI | 2 | 7.6 | 13/21 | 47.5s |
| #43 | MiMo-V2.5-Pro medium | Xiaomi | 2 | 7.5 | 12/21 | 26.1s |
| #58 | Gemini 3.1 Flash Lite Preview none | 2 | 7.2 | 12/21 | 1.21s | |
| #60 | Kimi K2.6 medium | Moonshot AI | 2 | 7.2 | 12/21 | 71.7s |
| #63 | GPT-5.3 Chat none | OpenAI | 2 | 7.2 | 12/21 | 6.34s |