Kategoria ya AI BENCHY
Orodha ya Maarifa ya jumla
Ona ni modeli gani za AI zinafanya vizuri zaidi katika Maarifa ya jumla, zipi zinabaki thabiti, na pengo kubwa liko wapi. Panga kwa: Muda wa majibu (wastani) ↓.
Sababu za kushindwa
| Nafasi | Modeli | Kampuni | Alama ya Maarifa ya jumla | Alama | Majaribio sahihi | Muda wa majibu (wastani) |
|---|---|---|---|---|---|---|
| #72 | GPT-5.5 none | OpenAI | 3.0 | 6.7 | 0/1 | 5.01s |
| #53 | GPT-5.4 Nano medium | OpenAI | 3.0 | 7.3 | 0/1 | 4.81s |
| #61 | Claude Sonnet 4.6 none | Anthropic | 3.0 | 7.2 | 0/1 | 4.67s |
| #46 | GPT-5.3 Chat none | OpenAI | 3.0 | 7.6 | 0/1 | 4.38s |
| #94 | Qwen3.6 27B none | Qwen | 3.0 | 5.8 | 0/1 | 4.03s |
| #112 | Kimi K2.5 none | Moonshot AI | 3.0 | 5.4 | 0/1 | 3.90s |
| #128 | MiMo-V2.5 none | Xiaomi | 3.0 | 4.9 | 0/1 | 3.89s |
| #78 | GLM 5 none | Z.ai | 3.0 | 6.5 | 0/1 | 3.62s |
| #27 | Gemini 3.1 Flash Lite medium | 3.0 | 7.9 | 0/1 | 3.08s | |
| #120 | DeepSeek V4 Flash none | DeepSeek | 3.0 | 5.2 | 0/1 | 3.07s |
| #25 | Gemini 2.5 Flash medium | 3.0 | 7.9 | 0/1 | 2.76s | |
| #7 | Gemini 3 Flash Preview low | 10.0 | 8.8 | 1/1 | 2.75s | |
| #133 | HY3 Preview none | Tencent | 3.0 | 4.6 | 0/1 | 2.71s |
| #23 | Gemini 3.1 Flash Lite Preview medium | 3.0 | 8.0 | 0/1 | 2.68s | |
| #80 | Mercury 2 medium | Inception | 3.0 | 6.3 | 0/1 | 2.58s |