Multi-Modal Foundation · Aignostics
RudolfV 2
A 1.1-billion-parameter pathology foundation model for image feature extraction in digital histopathology.
What RudolfV 2 is
RudolfV 2 is a pathology foundation model developed by Aignostics for computational pathology, biotechnology, and medical research. It processes histopathology images to generate visual representations for downstream analysis.
The model addresses the need for openly accessible pathology representations, providing an open foundation system developed to capture tissue morphology across biomedical imaging tasks.
How it works
RudolfV 2 is built on a Vision Transformer giant (ViT-g) architecture with approximately 1.1 billion parameters. Its self-supervised training procedure includes a dedicated post-training stage applied to the pathology representations.
The model has also been distilled into lightweight alternatives, including RudolfV 2-B, which employs a Vision Transformer base (ViT-B) architecture with 86 million parameters.
Key facts
- Architecture
- Vision Transformer giant (ViT-g)
- Parameter count
- Approximately 1.1 billion
- Core task
- Image feature extraction
- Distilled variants
- RudolfV 2-B (ViT-B, 86M parameters)
What you can use it for
- Histopathology feature extraction
- Generating numerical feature embeddings from tissue histology images for downstream analytical models.
- Computational pathology analysis
- Serving as a general foundation model for medical and biotechnology research investigating tissue specimens.
- Resource-adapted deployment
- Utilising distilled variants such as the 86-million-parameter RudolfV 2-B for applications with lower computational capacity.
Run it here
We list this model but cannot run it yet: gated on the Hub — never run past a gate the user did not accept.
Everything on this page still applies. Running it needs a loader for this repository, not a bigger machine.
Strengths and limits
Good at
- Offers high-capacity representation using a 1.1-billion-parameter Vision Transformer giant architecture.
- Incorporates a dedicated post-training stage specific to pathology foundation modelling.
- Provides smaller distilled versions, including an 86-million-parameter ViT-B model.
Watch out for
- The 1.1-billion-parameter size requires significant computational resources to run at scale.
- Implementation requires custom code when used through the transformers library.
Inputs and outputs
You provide
Image
PNG, JPEG or TIFF. Sent to the model as-is.
You get back
Mean-pooled representation, one vector per input.
Licence
Check the licence before commercial use · other
This model is gated by its authors — its licence has to be accepted on the source repository before it will run.
Questions
Can I use RudolfV 2 commercially?
Check the licence before commercial use. This model is gated by its authors — its licence has to be accepted on the source repository before it will run.
What architecture does RudolfV 2 use?
The model uses a Vision Transformer giant (ViT-g) architecture containing roughly 1.1 billion parameters.
What is the primary task of RudolfV 2?
Its primary function is image feature extraction for histopathology and computational pathology images.
Are there smaller versions of this model?
Yes, it has been distilled into smaller variants, including RudolfV 2-B, which uses a ViT-B architecture with 86 million parameters.
What training approach was used for RudolfV 2?
It uses self-supervised learning alongside a dedicated post-training stage.
Written from RudolfV 2's own documentation (model card). Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.