List your company
ResearchFormerly WizWand

Multimodal Reasoning on MMMUPro (Pass@1)

43.47Pass@1Self-distill Masking-KD-8B
8.900417.875226.8535.8248May 12, 2026
Updated 6mo ago

Evaluation results

MethodLinks
Self-distill Masking-KD-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=4096, Distillation=Self-distill2026.0543.47
Masking-KD-4BModel Size=~4B, Decoding Strategy=greedy, Max Length=4096, Distillation=Distilled from 8B2026.0540.52
Qwen3-VL-8B-ThinkingModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.0539.83
InternVL3.5-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.0538.5
InternVL3-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.0536.01
MiMo-VL-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.0534.57
Qwen3-VL-4B-ThinkingModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.0532.08
Qwen2.5-VL-7BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.0531.56
Masking-KD-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=4096, Distillation=Distilled from 8B2026.0530.75
InternVL3.5-4BModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.0529.77
Qwen2.5-VL-3BModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.0527.75
InternVL3-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.0522.95
Ovis2-8BModel Size=~8B, Decoding Strategy=greedy, Max Length=40962026.0516.18
Qwen3-VL-2B-ThinkingModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.0514.51
Ovis2-4BModel Size=~4B, Decoding Strategy=greedy, Max Length=40962026.0512.6
InternVL3.5-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.0512.54
Ovis2-2BModel Size=~2B, Decoding Strategy=greedy, Max Length=40962026.0510.23