Chinese Language Understanding on CLUEWSC
95.8Exact Match (EM)Opus-4.6
Evaluation Results
| Method | Links | |
|---|---|---|
| Opus-4.6Parameter Count=≈800B2026.06 | 95.8 | |
| Qwen3.5-PlusParameter Count=397B2026.06 | 95.6 | |
| GPT-5.4Parameter Count=≈1.5T2026.06 | 95.1 | |
| Qwen3-VL-32BDecoding Mode=Thinking, Model Scale=32B2026.06 | 94.9 | |
| Qwen3-VL-32BDecoding Mode=Instruct, Model Scale=32B2026.06 | 94.6 | |
| Yuvion VL-32BDecoding Mode=Reasoning, Model Scale=32B2026.06 | 93.3 | |
| DeepSeek-R1Architecture=MoE, Activated Params=37B, Total Params=671B2025.01 | 92.8 | |
| K2.5Parameter Count=≈1T2026.06 | 92.8 | |
| Yuvion VL-32BDecoding Mode=Instruct, Model Scale=32B2026.06 | 92.7 | |
| Qwen3-VL-8BDecoding Mode=Thinking, Model Scale=8B2026.06 | 92 | |
| Qwen3-VL-8BDecoding Mode=Instruct, Model Scale=8B2026.06 | 91.6 | |
| Yuvion VL-8BDecoding Mode=Reasoning, Model Scale=8B2026.06 | 91.1 | |
| DeepSeek-V3Architecture=MoE, Activated Params=37B, Total Params=671B2025.01 | 90.9 | |
| OpenAI-o1-mini2025.01 | 89.9 | |
| Yuvion VL-8BDecoding Mode=Instruct, Model Scale=8B2026.06 | 88.3 | |
| GPT-4o-05132025.01 | 87.9 | |
| Claude-3.5-Sonnet-10222025.01 | 85.4 |