Accelerating AI development in healthcare
AI building blocks for creating next-gen healthcare solutions.
Health AI Developer Foundations (HAI-DEF) is a collection of open-weight models and companion resources to help developers build AI models for healthcare.
Pre-trained models to jumpstart development
Access pre-trained models to build AI applications, with less data and compute resources.
Customizable for any application
HAI-DEF model weights are open, so you can run them wherever you'd like and fine-tune them with your data for your tasks.
Integrated with Google Cloud for scaled use
HAI-DEF models can be deployed on GCP as productionized, scalable services that natively integrate with Vertex AI, Google Cloud Storage, and Google Cloud DICOM Store.
Build with domain-specific models
Each HAI-DEF model is specialized for a specific medical domain. The models are developed using large amounts of diverse data for their respective domains, enabling application development that requires significantly less data and compute than traditional methods.
CXR Foundation
Accelerator for chest X-ray image analysis. Produces language-aligned embeddings that capture dense features relevant for chest X-ray applications.
- Data-efficient classification
- Zero-shot classification
- Semantic image retrieval
Path Foundation
Accelerator for histopathology image analysis. The model produces embeddings that capture dense features relevant for histopathology applications.
- Data-efficient classification
- Similar image search
Derm Foundation
Accelerator for skin image analysis. The model produces embeddings that capture dense features relevant for dermatology applications.
- Data-efficient classification
HeAR
Accelerator for audio analysis. The model produces embeddings that capture dense features relevant for audio applications.
- Data-efficient classification
- Data-efficient regression
Collection of open models to accelerate the development of therapeutics. TxGemma can make predictions about small molecules, proteins, nucleic acids, diseases, and cell lines. Can be used:
- for therapeutic classification, regression, and generation tasks
- conversationally through the chat variants
- for agentic orchestration
More models coming soon...
Access HAI-DEF models
Send feedback to: hai-def@google.com
We'd love to hear from you