Olympic medals are considered as a significant indicator of the success of a country at the international level. Forecasting the number of Olympic medals is critical for financial sponsors, media, politicians, and especially sports managers to evaluate team performance and strengthen success factors. Hence, the main aim of this paper is to evaluate the performance of the Artificial Neural Network (ANN) and Convolutional Neural Network (CNN) models, for accurate prediction of the number of medals of the USA, Italy, France, and Australia countries in Olympic games. Results indicated that the CNN model with the higher values for the correlation coefficient (0.980, 0.772, 0.857, and 0.932 for the USA, Italy, France, and Australia, respectively), lower values for the Root Mean Square Error, and less distance from observation points (10.155, 8.556, 7.435, and 11.485 for the USA, Italy, France, and Australia, respectively) is the most accurate model for forecasting the Olympic medals. Hence, based on the high importance of the most correct forecasts in the Olympic games, it is recommended to implement the CNN model. The forecast based on the CNN model cleared that the USA, Italy, France, and Australia will be earned 124, 51, 57, and 50 medals in Olympic 2024- Paris.