Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multimodal Evaluation Summary on Aggregated non-MME

84.4Average Accuracy

Qwen3-VL-32B

69.621673.458377.29581.1317May 9, 2026
Updated 15d ago

Evaluation Results

MethodLinks
2026.05
84.4100
2026.05
83.398.69
2026.05
82.5797.83
2026.05
81.9297.07
2026.05
81.6196.7
2026.05
81.4496.49
2026.05
81.4396.49
2026.05
8195.98
2026.05
79.8694.62
2026.05
79.1393.76
2026.05
79.05100
2026.05
78.76100
2026.05
78.4592.95
2026.05
77.8892.27
2026.05
77.6198.17
2026.05
77.3298.18
2026.05
77.2897.76
2026.05
76.9997.77
2026.05
76.8897.64
2026.05
76.8591.06
2026.05
76.7897.12
2026.05
76.7590.94
2026.05
76.6396.94
2026.05
76.6196.9
2026.05
76.5997.25
2026.05
76.2996.51
2026.05
76.2896.89
2026.05
76.1496.67
2026.05
76.0396.17
2026.05
75.9396.4
2026.05
75.8996.36
2026.05
75.5695.94
2026.05
75.5195.51
2026.05
75.2895.23
2026.05
75.2295.15
2026.05
75.1489.04
2026.05
74.9995.21
2026.05
73.9793.57
2026.05
73.4887.06
2026.05
73.4293.03
2026.05
73.392.99
2026.05
73.2292.9
2026.05
7392.61
2026.05
72.8692.16
2026.05
72.3391.5
2026.05
72.1191.21
2026.05
70.5589.48
2026.05
70.1988.79