List your company

Multimodal Understanding on MMMU Pro

55.7Vis AccuracyMiMo-VL-Miloco-7B
10.397622.158833.9245.6812Dec 17, 2025Dec 24, 2025Dec 31, 2025Jan 7, 2026Jan 14, 2026Jan 21, 2026Jan 28, 2026

Evaluation results

MethodLinks
MiMo-VL-Miloco-7B2025.1255.7
MiMo-VL-Miloco-7B2025.1247.2
InternVL3-8B2025.1245.6
MiMo-VL-7B-SFT-25082025.1243
Gemma-3-27B-IT2025.1237.8
InternVL3-8B2025.1237.8
MiMo-VL-7B-SFT-25082025.1236.4
Qwen2.5-VL-7B2025.1234.7
Qwen2.5VLSize=7B, Type=AR, Samples=>9M2025.1233.4
Qwen2.5-VL-7B2025.1229.4
DiffusionVLSize=7B, Type=Diff., Samples=738K2025.1225
Gemma-3-27B-IT2025.1224.9
Qwen2.5VLSize=3B, Type=AR, Samples=>9M2025.1222.1
DiffusionVLSize=3B, Type=Diff., Samples=738K2025.1220.2
LLaDA-VSize=8B, Type=Diff., Samples=16.5M2025.1218.6
LLaDA-VSettings=K=-, P=0, Avg (%)=100.0%, FLOPs (Rem./Red.)=100%2026.0118.38
LLaDA-FastVSettings=K=15, P=50, Avg (%)=98.7%, FLOPs (Rem./Red.)=51% (49%↓)2026.0117.57
LLaDA-FastVSettings=K=15, P=70, Avg (%)=95.3%, FLOPs (Rem./Red.)=37% (63%↓)2026.0116.47
Raw-FastVSettings=K=3, P=50, Avg (%)=80.0%, FLOPs (Rem./Red.)=37% (63%↓)2026.0113.64
Raw-FastVSettings=K=3, P=70, Avg (%)=68.8%, FLOPs (Rem./Red.)=21% (79%↓)2026.0112.14