Models Generic DINOv3 ViT-L/16
Meta

DINOv3 ViT-L/16

Latest iteration of Meta's self-supervised vision model.

Meta Open Source Vision Only
Generic
Model Type
#5
Overall Rank
40.2%
Avg P@1
29.8%
Avg mAP@10
1024
Embed Dim
224
Input Res
8
Datasets

Performance Across Datasets

Dataset Category P@1 P@5 R@1 R@5 mAP@10
VPRC 2023 Mixed Retail 24.03% 10.74% 15.51% 31.87% 25.57%
Intercars Automotive 26.29% 21.23% 8.82% 24.78% 26.40%
Stanford Online Products E-commerce 48.00% 25.90% 10.88% 24.48% 24.89%
IKEA Furniture 48.52% 28.52% 15.86% 33.72% 32.16%
Hornbach Hardware/DIY 24.22% 9.16% 24.22% 45.81% 31.96%
ARaymond Industrial 26.38% 16.87% 1.65% 5.27% 8.23%
Products-10K E-commerce 59.13% 37.13% 12.54% 38.71% 37.70%
TOPEX Industrial 65.25% 61.08% 2.04% 9.54% 51.47%
ILIAS Instance Retrieval 47.97% 20.91% 11.31% 22.47% 21.67%
Average 40.23% 26.33% 11.44% 26.77% 29.80%