<p>This study addresses the escalating impact of climate change, leading to an increased frequency and severity of global drought occurrences. The mitigation of potential drought impacts necessitates the implementation of effective water management practices. Within this framework, drought forecasting assumes a significant role as a crucial tool supporting socio-economic initiatives. Precise prediction of forthcoming drought events holds practical implications for risk management, proactive drought preparedness, and efficient farm irrigation. This study evaluates the performance of various predictive models, autoregressive integrated moving average (ARIMA), support vector machine (SVM), artificial neural networks (ANN), long short-term memory (LSTM), and a genetic algorithm (GA) optimized hybrid model, in forecasting standardized precipitation index (SPI) obtained from monthly precipitation data from 1975 to 2022 for Jhansi and Banda districts. The model performance was assessed using root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). Results show that the LSTM model outperformed traditional models (ARIMA) and other machine learning models (SVM, ANN), demonstrating superior handling of time series data. However, the GA Hybrid model, which combines and optimizes the strengths of individual models, yielded the most accurate predictions by reducing RMSE, MAE and MAPE.</p>