Kushindwa kwa AI BENCHY
Kushindwa kwa Hakufuata maelekezo
Ona ni modeli gani za AI hukutana na Hakufuata maelekezo mara nyingi zaidi ili utambue hatari za utegemevu kabla ya kuchagua. Panga kwa: Idadi ya kushindwa ↑.
| Nafasi | Modeli | Kampuni | Idadi ya Hakufuata maelekezo | Alama | Majaribio sahihi | Muda wa majibu (wastani) |
|---|---|---|---|---|---|---|
| #9 | Qwen3.6 Plus Preview medium | Qwen | 1 | 8.5 | 13/17 | 13.9s |
| #11 | Gemini 3.1 Flash Lite Preview high | 1 | 8.4 | 12/16 | 68.8s | |
| #13 | GLM 5 medium | Z.ai | 1 | 8.4 | 13/18 | 23.3s |
| #14 | Gemma 4 31B medium | 1 | 8.3 | 13/18 | 24.9s | |
| #15 | Gemini 2.5 Flash medium | 1 | 8.2 | 13/18 | 12.1s | |
| #17 | Gemini 3.1 Flash Lite Preview medium | 1 | 8.2 | 13/18 | 3.74s | |
| #20 | Qwen3.6 Plus medium | Qwen | 1 | 8.1 | 13/18 | 15.3s |
| #22 | Gemini 3.1 Flash Lite Preview low | 1 | 8.1 | 13/18 | 3.22s | |
| #23 | MiMo-V2-Pro medium | Xiaomi | 1 | 8.1 | 12/18 | 12.3s |
| #24 | Gemma 4 26B A4B medium | 1 | 8.0 | 13/18 | 25.0s | |
| #27 | DeepSeek V3.2 medium | DeepSeek | 1 | 8.0 | 12/18 | 46.4s |
| #28 | GPT-5.2 Chat none | OpenAI | 1 | 7.9 | 12/18 | 6.84s |
| #32 | Qwen3.5-Flash medium | Qwen | 1 | 7.8 | 11/18 | 66.7s |
| #39 | Seed-2.0-Mini medium | Bytedance Seed | 1 | 7.5 | 11/18 | 69.7s |
| #41 | MiMo-V2-Flash medium | Xiaomi | 1 | 7.5 | 11/18 | 23.4s |