Visual Understanding on SAT
73.3AccuracyGPT-5
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5Size=-2025.10 | 73.3 | |
| Qwen2.5-VLSize=72B2025.10 | 64.67 | |
| Gemini 2.5 FlashSize=-2025.10 | 56 | |
| ThinkMorphSize=7B, think mode=true2025.10 | 52.67 | |
| Qwen2.5-VLSize=7B2025.10 | 51.33 | |
| InternVL3.5Size=38B2025.10 | 49.33 | |
| InternVL3.5Size=8B2025.10 | 45.33 | |
| BagelSize=7B, think mode=true2025.10 | 44.67 | |
| GPT-4oSize=-2025.10 | 28 | |
| Janus-proSize=7B2025.10 | 22 | |
| ChameleonSize=7B2025.10 | 10.67 |