Artificial intelligence (AI) greatly improves the accuracy and efficiency of weather radar casting, which is currently essential for short-term weather forecasts. To enhance real-time weather forecasting, this study investigates AI-driven nowcasting models, such as CNN-LSTM and Transformer networks. Using measures like MSE, CSI, and POD, we evaluate the efficacy of AI-based techniques against conventional numerical weather prediction (NWP) and radar extrapolation. The outcomes show that AI is better at forecasting severe weather conditions like thunderstorms and cyclones. With its increased accuracy, speed, and adaptability, AI-driven casting is now a crucial tool for disaster preparedness and meteorological applications.