Multimodal Understanding on VLM Evaluation Suite (Hall, MME, AI2D, RWQA, SQA, POPE, MBen, MBzh, CCB, VSR, V7W)
63.76HallQwen3-VL-32B
Evaluation results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen3-VL-32BRatio=100%, Backbone=Qwen3-VL-32B2026.05 | 63.76 | 2,501.3 | 87.63 | 75.82 | 96.72 | 89.11 | 87.64 | 86.3 | 78.13 | 88.87 | 90 | 84.4 | 100 | |
| F3ARatio=60%, Backbone=Qwen3-VL-32B2026.05 | 62.5 | 2,494.3 | 84.84 | 74.64 | 94.97 | 88.89 | 86.46 | 85.35 | 78.22 | 88.22 | 88.86 | 83.3 | 98.69 | |
| FastVRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 62.19 | 2,311.4 | 81.51 | 70.07 | 93.18 | 89.46 | 86.07 | 85.37 | 78.45 | 87.4 | 87.9 | 82.16 | 98.8 | |
| DivPruneRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 62.19 | 2,341 | 82.71 | 70.2 | 93.18 | 89.96 | 85.91 | 85.3 | 78.18 | 85.76 | 88.27 | 82.17 | 98.8 | |
| FastVRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 62.19 | 2,311.4 | 81.51 | 70.07 | 93.18 | 89.46 | 86.07 | 85.37 | 78.45 | 87.4 | 87.9 | 82.16 | 98.8 | |
| DivPruneRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 62.19 | 2,341 | 82.71 | 70.2 | 93.18 | 89.96 | 85.91 | 85.3 | 78.18 | 85.76 | 88.27 | 82.17 | 98.8 | |
| Qwen3-VL-30B-A3BRatio=100%, Backbone=Qwen3-VL-30B-A3B2026.05 | 61.87 | 2,381.7 | 84.55 | 71.11 | 95.13 | 90.2 | 86.74 | 85.79 | 79.01 | 87.15 | 90.06 | 83.16 | 100 | |
| Qwen3-VL-30B-A3BRatio=100%, Backbone=Qwen3-VL-30B-A3B2026.05 | 61.87 | 2,381.7 | 84.55 | 71.11 | 95.13 | 90.2 | 86.74 | 85.79 | 79.01 | 87.15 | 90.06 | 83.16 | 100 | |
| VisionZipRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 61.66 | 2,285.6 | 82.74 | 69.93 | 93.79 | 90.31 | 86.46 | 85.56 | 78.31 | 86.82 | 88.1 | 82.37 | 99.05 | |
| VisionZipRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 61.66 | 2,285.6 | 82.74 | 69.93 | 93.79 | 90.31 | 86.46 | 85.56 | 78.31 | 86.82 | 88.1 | 82.37 | 99.05 | |
| F3ARatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 61.13 | 2,375.2 | 82.48 | 70.46 | 93.33 | 90.3 | 85.95 | 85.95 | 78.55 | 86.42 | 89.48 | 82.41 | 99.09 | |
| F3ARatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 61.13 | 2,375.2 | 82.48 | 70.46 | 93.33 | 90.3 | 85.95 | 85.95 | 78.55 | 86.42 | 89.48 | 82.41 | 99.09 | |
| F3ARatio=40%, Backbone=Qwen3-VL-32B2026.05 | 61.03 | 2,355.6 | 82.55 | 73.33 | 92.1 | 88.8 | 84.98 | 84.73 | 77.3 | 86.58 | 87.82 | 81.92 | 97.07 | |
| VisionZipRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 60.71 | 2,295.8 | 80.8 | 69.54 | 92.05 | 90.04 | 85.17 | 84.06 | 78.64 | 86.33 | 86.63 | 81.4 | 97.88 | |
| VisionZipRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 60.71 | 2,295.8 | 80.8 | 69.54 | 92.05 | 90.04 | 85.17 | 84.06 | 78.64 | 86.33 | 86.63 | 81.4 | 97.88 | |
| DivPruneRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 60.5 | 2,277.8 | 80.86 | 69.28 | 92.1 | 89.44 | 85.03 | 84.33 | 77.95 | 85.43 | 85.96 | 81.09 | 97.51 | |
| F3ARatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 60.5 | 2,315.2 | 82.19 | 69.93 | 91.79 | 90.02 | 85.7 | 84.82 | 78.27 | 86.25 | 88.96 | 81.84 | 97.88 | |
| DivPruneRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 60.5 | 2,277.8 | 80.86 | 69.28 | 92.1 | 89.44 | 85.03 | 84.33 | 77.95 | 85.43 | 85.96 | 81.09 | 97.51 | |
| F3ARatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 60.5 | 2,315.2 | 82.19 | 69.93 | 91.79 | 90.02 | 85.7 | 84.82 | 78.27 | 86.25 | 88.96 | 81.84 | 97.88 | |
| FastVRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 59.98 | 2,235.2 | 79.63 | 68.5 | 91.48 | 87.68 | 85.1 | 84.52 | 77.95 | 85.6 | 85.48 | 80.59 | 96.91 | |
| FastVRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 59.98 | 2,235.2 | 79.63 | 68.5 | 91.48 | 87.68 | 85.1 | 84.52 | 77.95 | 85.6 | 85.48 | 80.59 | 96.91 | |
| CDPrunerRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 59.35 | 2,364.3 | 81.35 | 69.8 | 92.41 | 89.88 | 85.65 | 85.33 | 77.67 | 86.09 | 88.29 | 81.58 | 98.1 | |
| CDPrunerRatio=60%, Backbone=Qwen3-VL-30B-A3B2026.05 | 59.35 | 2,364.3 | 81.35 | 69.8 | 92.41 | 89.88 | 85.65 | 85.33 | 77.67 | 86.09 | 88.29 | 81.58 | 98.1 | |
| CDPrunerRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 58.61 | 2,211.8 | 78.85 | 69.54 | 90.1 | 88.63 | 84.63 | 83.27 | 77.85 | 85.43 | 87.4 | 80.43 | 96.72 | |
| CDPrunerRatio=40%, Backbone=Qwen3-VL-30B-A3B2026.05 | 58.61 | 2,211.8 | 78.85 | 69.54 | 90.1 | 88.63 | 84.63 | 83.27 | 77.85 | 85.43 | 87.4 | 80.43 | 96.72 | |
| F3ARatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 57.98 | 2,189.7 | 78.3 | 67.58 | 89.43 | 89.14 | 84.47 | 83.66 | 76.61 | 82.9 | 87.3 | 79.74 | 95.88 | |
| F3ARatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 57.98 | 2,189.7 | 78.3 | 67.58 | 89.43 | 89.14 | 84.47 | 83.66 | 76.61 | 82.9 | 87.3 | 79.74 | 95.88 | |
| F3ARatio=20%, Backbone=Qwen3-VL-32B2026.05 | 56.61 | 2,269.2 | 77.56 | 61.83 | 86.45 | 87.1 | 83.06 | 82.16 | 75.09 | 84.04 | 84.88 | 77.88 | 92.27 | |
| FastVRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 56.51 | 1,997 | 74.32 | 63.53 | 89.79 | 83.32 | 82.51 | 82.26 | 74.95 | 82.24 | 80.12 | 76.95 | 92.54 | |
| FastVRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 56.51 | 1,997 | 74.32 | 63.53 | 89.79 | 83.32 | 82.51 | 82.26 | 74.95 | 82.24 | 80.12 | 76.95 | 92.54 | |
| DivPruneRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 56.4 | 2,143.3 | 78.14 | 65.88 | 88.76 | 86.9 | 82.72 | 82.28 | 77.53 | 82.65 | 81.17 | 78.24 | 94.09 | |
| DivPruneRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 56.4 | 2,143.3 | 78.14 | 65.88 | 88.76 | 86.9 | 82.72 | 82.28 | 77.53 | 82.65 | 81.17 | 78.24 | 94.09 | |
| VisionZipRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 55.67 | 1,972.6 | 74 | 67.06 | 87.48 | 87.97 | 82.37 | 80.66 | 75.14 | 82.41 | 82.36 | 77.51 | 93.21 | |
| VisionZipRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 55.67 | 1,972.6 | 74 | 67.06 | 87.48 | 87.97 | 82.37 | 80.66 | 75.14 | 82.41 | 82.36 | 77.51 | 93.21 | |
| F3ARatio=60%, Backbone=Qwen2.5-VL-7B2026.05 | 54.62 | 2,213.9 | 79.57 | 62.61 | 86.81 | 86.3 | 81.16 | 81.82 | 73.88 | 81.25 | 85.16 | 77.32 | 98.18 | |
| Qwen2.5-VL-7BRatio=100%, Backbone=Qwen2.5-VL-7B2026.05 | 53.41 | 2,251 | 82.25 | 67.32 | 89.12 | 87.59 | 82.1 | 82.37 | 75.01 | 81.66 | 86.72 | 78.76 | 100 | |
| F3ARatio=40%, Backbone=Qwen2.5-VL-7B2026.05 | 53.35 | 2,166.5 | 76.85 | 61.05 | 84.76 | 84.56 | 80.52 | 81.94 | 73.51 | 80.7 | 84.12 | 76.14 | 96.67 | |
| CDPrunerRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 52.83 | 2,049.8 | 74.9 | 62.48 | 87.38 | 87.2 | 83.02 | 81.15 | 74.82 | 82.32 | 85.09 | 77.12 | 92.73 | |
| CDPrunerRatio=20%, Backbone=Qwen3-VL-30B-A3B2026.05 | 52.83 | 2,049.8 | 74.9 | 62.48 | 87.38 | 87.2 | 83.02 | 81.15 | 74.82 | 82.32 | 85.09 | 77.12 | 92.73 | |
| F3ARatio=20%, Backbone=Qwen2.5-VL-7B2026.05 | 49.57 | 1,989 | 70.95 | 56.86 | 82.91 | 82.38 | 81 | 80.24 | 69.02 | 78.48 | 82.82 | 73.42 | 93.03 |