Pathology Models
Computational pathology models apply vision transformers and self-supervised learning to whole-slide histology images, enabling automated analysis of tissue architecture, cell morphology, and disease-relevant spatial patterns at a scale and consistency no human pathologist can match. These models power applications including cancer grading, biomarker prediction from H&E stains, and survival analysis, reducing the bottleneck of expert annotation in clinical and research workflows. By learning rich visual representations from millions of pathology images, they are beginning to reveal morphological features that correlate with molecular subtypes and treatment response.
5 models in this category
Notable Models
Top-rated pathology models from our evaluations
H-optimus-0
Bioptimus
A 1.1B parameter open-source vision transformer for histopathology, trained on 500,000+ H&E whole slide images from 4,000 clinical practices worldwide.
Hibou
HistAI
DINOv2-based Vision Transformer foundation models for digital pathology, trained on over 1 million whole-slide images. Available as Hibou-B (86M) and Hibou-L (307M) under Apache 2.0.
Prov-GigaPath
Microsoft Research
Whole-slide pathology foundation model pretrained on 1.3 billion tiles from 171,189 clinical WSIs. Achieves state-of-the-art on 25 of 26 pathology benchmark tasks.
UNI
Mahmood Lab
Self-supervised pathology foundation model (ViT-L/16, DINOv2) pretrained on 100M+ H&E tiles from 100,000+ whole-slide images. State-of-the-art on 34 pathology tasks.
Virchow
Paige AI
Self-supervised vision transformer foundation models for computational pathology, pre-trained on up to 3.1 million whole slide images from 632M to 1.9B parameters.