TabPFN-3.5

TabPFN-3.5 is a machine learning model designed to make state-of-the-art predictions on complex, messy tabular datasets commonly found in business and scientific applications. It enables zero-shot inference, outperforms traditional tuned models, and works well on datasets with text-rich features, temporal and grouped data without extensive feature engineering. Available via API, open-source package, and integration with cloud platforms.