Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Question Answering on MMBench (dev)
Loading...
77.3
Accuracy
Qwen
55.356
61.053
66.75
72.447
Aug 1, 2025
Accuracy
Average Relative Performance
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Average Relative Performance
Qwen
Backbone=Qwen2.5-VL-3B...
2025.08
77.3
100
HiPrune++
Backbone=Qwen2.5-VL-3B...
2025.08
76
98.7
HiPrune
Backbone=Qwen2.5-VL-3B...
2025.08
75.9
98.7
VisionZip
Backbone=Qwen2.5-VL-3B...
2025.08
74.9
97.7
FastV
Backbone=Qwen2.5-VL-3B...
2025.08
74.4
97.4
HiPrune++
Backbone=Qwen2.5-VL-3B...
2025.08
74.2
97.1
HiPrune
Backbone=Qwen2.5-VL-3B...
2025.08
73.7
97.1
VisionZip
Backbone=Qwen2.5-VL-3B...
2025.08
73.5
96.2
FastV
Backbone=Qwen2.5-VL-3B...
2025.08
72.4
95.9
HiPrune++
Backbone=Qwen2.5-VL-3B...
2025.08
70.5
93
HiPrune
Backbone=Qwen2.5-VL-3B...
2025.08
69.6
93
VisionZip
Backbone=Qwen2.5-VL-3B...
2025.08
67.8
91.5
FastV
Backbone=Qwen2.5-VL-3B...
2025.08
56.2
86.4
Feedback
Search any
task
Search any
task
