| 🇯🇵 |
1 |
sakana: namazu |
90.33 |
| 🇨🇳 |
2 |
Qwen: qwen3.8-27b通义千问 |
90.3 |
| 🇨🇳 |
3 |
Moonshot AI: Kimi K2.6月之暗面 |
89.6 |
| 🇺🇸 |
4 |
NVIDIA Corporation: nvidia-nemotron-3.5-super-vl-120b-a12b |
88.22 |
| 🇺🇸 |
5 |
Anthropic: claude-opus-4.5 |
84.8 |
| ❓ |
6 |
Orion LLM Labs: grm-2.6-plus |
84.8 |
| 🇨🇳 |
7 |
Qwen: qwen3.6-27b通义千问 |
83.9 |
| ❓ |
8 |
krafton: raon-1-mini-preview |
81.6 |
| 🇨🇳 |
9 |
Qwen: Qwen3.5 27B通义千问 |
80.7 |
| 🇰🇷 |
10 |
LG AI Research: k-exaone-236b-a23b주식회사 엘지 |
80.7 |
| ❓ |
11 |
recursivesi: flint |
80.3 |
| ❓ |
12 |
weiboai: vibethinker-3b |
80.2 |
| 🇺🇸 |
13 |
Google: Gemma 4 31B Instruct |
80 |
| ❓ |
14 |
diffuspace: clover-35b-a3b-preview |
72.6 |
| 🇺🇸 |
15 |
Google: Gemma 4 12B Instruct |
72 |
| 🇨🇦 |
16 |
Cohere Labs: north-mini-code-1.0 |
70.3 |
| 🇳🇱 |
17 |
JetBrains: Mellum2 Thinking |
69.9 |
| 🇨🇳 |
18 |
OpenBMB: minicpm5-2b |
69.1 |
| 🇺🇸 |
19 |
Google: diffusiongemma-26b-a4b-it |
69.1 |
| 🇺🇸 |
20 |
NVIDIA Corporation: nvidia-nemotron-3-nano-30b-a3b |
68.3 |
| ❓ |
21 |
zyphra: zaya1-8b |
65.8 |
| 🇨🇳 |
22 |
Qwen: Qwen3.5 9B通义千问 |
65.6 |
| 🇺🇸 |
23 |
Apple: simplesd-4b-thinking |
57.8 |
| 🇷🇺 |
24 |
Sber: gigachat3.5-432b-a28bСбер |
56.2 |
| 🇺🇸 |
25 |
Apple: simplesd-30b-instruct |
55.3 |
| 🇺🇸 |
26 |
Google: Gemma 4 E4B Instruct |
52 |
| 🇨🇳 |
27 |
Qwen: Qwen3 4B通义千问 |
48.5 |
| 🇨🇳 |
28 |
Tencent: hunyuan-4b-instruct腾讯控股有限公司 |
46.8 |
| 🇺🇸 |
29 |
OpenAI: gpt-oss-20b |
45.8 |
| 🇨🇳 |
30 |
Tencent: youtu-llm-2b腾讯控股有限公司 |
43.7 |
| 🇺🇸 |
31 |
Apple: simplesd-4b-instruct |
41.5 |
| 🇨🇳 |
32 |
Xiaomi: MiMo-V2.5 Pro小米集团 |
39.6 |
| ❓ |
33 |
essentialai: rnj-1-instruct |
35.2 |
| 🇨🇳 |
34 |
Tencent: hy3-preview-base腾讯控股有限公司 |
34.86 |
| 🇺🇸 |
35 |
Hugging Face Smol Models Research: smollm3-3b |
30.8 |
| 🇨🇳 |
36 |
Qwen: qwen3-1.7b通义千问 |
30.7 |
| 🇨🇳 |
37 |
DeepSeek: deepseek-r1-distill-qwen-1.5b深度求索 |
19.8 |