This paper presents a space–time landslide hazard modeling approach using an Ensemble Neural Network (ENN). The study aims to jointly model landslide occurrences and their associated areal density per mapping unit in space and time. The model is trained on a spatio-temporal landslide database generated for the Nepalese region affected by the Gorkha earthquake. The model uses a deep-learning architecture trained with an ENN, where landslide occurrences and densities are aggregated over a 1 km × 1 km squared mapping unit and classified or regressed against a nested 30 m lattice. The model incorporates predisposing and triggering factors at the nested level and uses an approximately 6-month temporal resolution. The results show that the model performs well in both susceptibility (AUC=0.93) and density prediction (Pearson r=0.93) tasks over the entire spatio-temporal domain. This model is the first to propose an integrated framework for hazard modeling in a data-driven context, addressing the three components of hazard definition: location, frequency, and size. The study also highlights the importance of considering both spatial and temporal aspects in landslide hazard modeling, and the model's results demonstrate its effectiveness in predicting landslide susceptibility and area density. The model's performance is evaluated using various metrics, including F1 score, Intersection over Union (IOU), and Pearson's R coefficient. The results show that the model achieves high accuracy in predicting landslide susceptibility and area density, with a strong correlation between observed and predicted values. The study also discusses the implications of the model's results for landslide hazard assessment and management, emphasizing the importance of considering both spatial and temporal dimensions in landslide hazard modeling. The model's results suggest that the hazard is primarily concentrated in areas with high susceptibility and high area density, and that the hazard decreases over time. The study concludes that the proposed ENN model is a promising approach for space–time landslide hazard modeling, providing a comprehensive evaluation of landslide occurrences and their size.This paper presents a space–time landslide hazard modeling approach using an Ensemble Neural Network (ENN). The study aims to jointly model landslide occurrences and their associated areal density per mapping unit in space and time. The model is trained on a spatio-temporal landslide database generated for the Nepalese region affected by the Gorkha earthquake. The model uses a deep-learning architecture trained with an ENN, where landslide occurrences and densities are aggregated over a 1 km × 1 km squared mapping unit and classified or regressed against a nested 30 m lattice. The model incorporates predisposing and triggering factors at the nested level and uses an approximately 6-month temporal resolution. The results show that the model performs well in both susceptibility (AUC=0.93) and density prediction (Pearson r=0.93) tasks over the entire spatio-temporal domain. This model is the first to propose an integrated framework for hazard modeling in a data-driven context, addressing the three components of hazard definition: location, frequency, and size. The study also highlights the importance of considering both spatial and temporal aspects in landslide hazard modeling, and the model's results demonstrate its effectiveness in predicting landslide susceptibility and area density. The model's performance is evaluated using various metrics, including F1 score, Intersection over Union (IOU), and Pearson's R coefficient. The results show that the model achieves high accuracy in predicting landslide susceptibility and area density, with a strong correlation between observed and predicted values. The study also discusses the implications of the model's results for landslide hazard assessment and management, emphasizing the importance of considering both spatial and temporal dimensions in landslide hazard modeling. The model's results suggest that the hazard is primarily concentrated in areas with high susceptibility and high area density, and that the hazard decreases over time. The study concludes that the proposed ENN model is a promising approach for space–time landslide hazard modeling, providing a comprehensive evaluation of landslide occurrences and their size.