Skip to content

Prov-GigaPath

A whole-slide foundation model for digital pathology

Visit official site
Category AI models

What Prov-GigaPath is

The model is described as a whole-slide foundation model for digital pathology research trained on real-world data and published in Nature.

What you can do with it

  • Encode local patterns with the tile encoder
  • Produce slide-level representations with the slide encoder
  • Fine-tune with pre-extracted PCam and PANDA embeddings

Application scenarios

  • A researcher analyses pathology data for a publication
  • A medical school teaches digital pathology
  • A hospital team automates slide analysis for research

What to watch out for

  • The model is stated for research use only
  • Clinical care and clinical decisions are excluded
  • HuggingFace Hub terms must be accepted for model access

Pros & cons

✓ What we like

  • Pretrained models, code and demo notebooks are provided
  • The model is available on HuggingFace Hub

! What to watch out for

  • The model is stated for research use only
  • Clinical care and clinical decisions are excluded

FAQ

Can it be used clinically?

The description states it is not for clinical care or clinical decision-making and is limited to research.

What hardware is named?

NVIDIA A100 Tensor Core GPU machines are listed.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review