Multimodal Integration on MMStar
66.49AccuracyOurs
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OursModel=Kimi-VL A3B-Thinking2025.10 | 66.49 | 49.78 | |
| AGLAModel=Kimi-VL A3B-Thinking2025.10 | 63.76 | 47.06 | |
| VanillaModel=Kimi-VL A3B-Thinking2025.10 | 59.76 | 42.75 | |
| VCDModel=Kimi-VL A3B-Thinking2025.10 | 59.48 | 42.61 | |
| OursModel=R1-Onevision 7B2025.10 | 58.02 | 40.94 | |
| VCDModel=R1-Onevision 7B2025.10 | 56.74 | 39.61 | |
| VanillaModel=R1-Onevision 7B2025.10 | 56.26 | 39.11 | |
| AGLAModel=R1-Onevision 7B2025.10 | 54.49 | 37.86 | |
| OursModel=Ocean-R1 7B Instruct2025.10 | 50.77 | 34.11 | |
| CGDModel=Kimi-VL A3B-Thinking2025.10 | 48.73 | 31.09 | |
| AGLAModel=Ocean-R1 7B Instruct2025.10 | 48.48 | 32.27 | |
| VCDModel=Ocean-R1 7B Instruct2025.10 | 45.48 | 29.23 | |
| VanillaModel=Ocean-R1 7B Instruct2025.10 | 45.24 | 29.38 | |
| CGDModel=R1-Onevision 7B2025.10 | 43.48 | 26.87 | |
| CGDModel=Ocean-R1 7B Instruct2025.10 | 36.24 | 21.51 |