DeepPurpose provides a comprehensive toolkit for molecular modeling using deep learning, supporting tasks such as drug-target interaction (DTI) prediction, drug repurposing, virtual screening, QSAR modeling, and protein function prediction. It offers various molecular encoding schemes including GNN-based models (DGL_GCN, DGL_GIN, AttentiveFP) and simplified dataset loading through integration with the Therapeutics Data Commons (TDC). The library enables researchers to build predictive models with just a few lines of code using PyTorch as the underlying framework.