MM Document Understanding on AI2D (test)
91.5ScoreQwen3.5-Plus
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3.5-PlusParameter Count=397B2026.06 | 91.5 | |
| K2.5Parameter Count=≈1T2026.06 | 90.5 | |
| GPT-5.4Parameter Count=≈1.5T2026.06 | 89.4 | |
| Qwen3-VL-32BDecoding Mode=Thinking, Model Scale=32B2026.06 | 88.2 | |
| Qwen3-VL-32BDecoding Mode=Instruct, Model Scale=32B2026.06 | 88.1 | |
| Opus-4.6Parameter Count=≈800B2026.06 | 87.2 | |
| Yuvion VL-32BDecoding Mode=Reasoning, Model Scale=32B2026.06 | 85.7 | |
| Qwen3-VL-8BDecoding Mode=Thinking, Model Scale=8B2026.06 | 84.8 | |
| Yuvion VL-8BDecoding Mode=Reasoning, Model Scale=8B2026.06 | 84.8 | |
| Yuvion VL-32BDecoding Mode=Instruct, Model Scale=32B2026.06 | 84.1 | |
| Qwen3-VL-8BDecoding Mode=Instruct, Model Scale=8B2026.06 | 83.9 | |
| Yuvion VL-8BDecoding Mode=Instruct, Model Scale=8B2026.06 | 81.6 |