Extreme climatic events like heat wave affect well-being of people which is central concern under the Sustainable Cities agenda initiated by United Nations. The increasing heat stress within urban areas not only poses risks to the health of the residents but also amplifies the energy requirements of the constructed environment. This research focuses on predicting heat stress using the Wet Bulb Globe Temperature (WBGT) through artificial neural network (ANN) models during typical summer months (March to June) in four streets of Gwalior, India. These models utilized air temperature, relative humidity, wet bulb temperature, the month of the year, day of the month, and the hour of the day as input parameters, with WBGT as the target parameter. The range of hidden neurons derived in the study allowed for development of forty neural network models. The most favorable results were achieved using configurations of 7 hidden neurons and 2 hidden layers. The models’ performance was assessed through the coefficient of correlation (R2) and root mean square error (RMSE). The R2 and RMSE values ranged from 0.97 to 0.99 and 0.88–1.89, respectively, for the four selected streets. These outcomes highlight the exceptional predictive capability of the developed ANN models for WBGT prediction. This study makes a noteworthy contribution by demonstrating the potential of ANN as a valuable tool for the Sustainable Cities program, particularly in forecasting urban heat stress. The study will contribute toward new methodology for heat stress forecasting and real-time policymaking for more sustainable and climate-resilient urban environments.