Deployment
Databricks
Predict using tabular foundation model directly on Unity Catalog–governed Delta tables or within MLflow. Best-in-class results without hyperparameter tuning or retraining cycles.
SOTA predictions
TabPFN exceeds the accuracy of traditional ML methods that require hours of tuning and expert oversight.
No preprocessing
No time spent on feature engineering and data cleaning. TabPFN handles Databricks warehouse data as-is.
Inside Databricks
Use TabPFN inside existing Databricks setup, maintaining data governance and control.
How it works
Choose how to get started
Try TabPFN with your Databricks setup
Use open-source solution cookbook to see the power of TabPFN on Databricks. Use MLflow, AI functions or Databricks notebooks.
Chat with Databricks expert
Not sure where to start? Chat with our Databricks expert to identify best use cases