Ludwig, an open-source declarative machine learning framework, has expanded its capabilities in version 0.6 by introducing gradient boosted tree (GBM) models for tabular data, complementing its existing support for neural network (NN) models. GBMs offer faster training speeds and often outperform NNs on small tabular datasets, particularly in cases with class imbalance, making them a valuable addition for users seeking a unified interface for model experimentation. While NNs remain versatile for handling various data types and multi-task learning, GBMs provide a competitive alternative for binary, categorical, or regression tasks with a single output feature. Ludwig's configuration allows easy switching between model types, enabling efficient comparisons and optimizations using its built-in preprocessing, hyperparameter tuning, and visualization tools. This enhancement positions Ludwig as a comprehensive platform for training diverse model types, and users are encouraged to participate in its growing community and explore enterprise solutions like Predibase for streamlined machine learning project management.