AI BENCHY Fouten
Instructies niet gevolgd-fouten
Zie welke AI-modellen het vaakst tegen Instructies niet gevolgd aanlopen, zodat je betrouwbaarheidsrisico's ziet voordat je kiest. Sorteren op: Responstijd (gem.) ↑.
| Rang | Model | Bedrijf | Instructies niet gevolgd-aantal | Score | Correcte tests | Responstijd (gem.) |
|---|---|---|---|---|---|---|
| #70 | GPT-5.4 Nano medium | OpenAI | 2 | 7.0 | 11/21 | 12.0s |
| #113 | DeepSeek V4 Pro none | DeepSeek | 2 | 5.7 | 7/21 | 12.4s |
| #156 | Hy3 preview none | Tencent | 4 | 4.4 | 4/21 | 12.9s |
| #124 | Kimi K2.6 none | Moonshot AI | 3 | 5.5 | 7/21 | 13.3s |
| #133 | DeepSeek V3.2 none | DeepSeek | 1 | 5.2 | 6/21 | 13.8s |
| #92 | Laguna M.1 medium | Poolside | 1 | 6.4 | 9/19 | 14.7s |
| #28 | Gemini 2.5 Flash medium | 1 | 7.8 | 14/21 | 15.5s | |
| #15 | GPT-5.3-Codex medium | OpenAI | 2 | 8.4 | 15/21 | 16.2s |
| #33 | Hy3 preview medium | Tencent | 1 | 7.7 | 14/21 | 16.3s |
| #42 | GPT-5.2 medium | OpenAI | 3 | 7.5 | 13/21 | 16.9s |
| #149 | Nemotron 3 Nano Omni 30b A3b Reasoning medium | NVIDIA | 1 | 4.6 | 4/19 | 17.1s |
| #46 | Qwen3.6 35B A3B medium | Qwen | 1 | 7.4 | 13/21 | 18.1s |
| #39 | Qwen3.6 Flash medium | Qwen | 1 | 7.5 | 12/21 | 19.2s |
| #64 | MiMo-V2-Flash medium | Xiaomi | 1 | 7.2 | 12/21 | 20.1s |
| #22 | Step 3.7 Flash medium | Stepfun | 1 | 8.0 | 14/21 | 20.4s |