Forecasting maximum daily electricity price (MDEP) is important due to its benefits for not only hydropower operators but also transmission system and distribution network operators. Based on the proven applicability of artificial intelligence tools for energy price prediction, this article aims to quantify the uncertainty of Artificial Neural Networks (ANNs) and hybrid convolution Neural Networks-Long Short-Term Memory (CNN-LSTM) models for MDEP forecasting using Monte Carlo simulation. The uncertainties were analyzed by calculating P-factor and d-factor indices and analyzing outliers and distribution of errors in different seasons. The results demonstrate the higher performance of the CNN-LSTM model compared to the ANN model. The CNN-LSTM has a d-factor less than 0.961 while the ANN model has a d-factor of 1.21 in spring season. The prediction performance of the CNN-LSTM model is strongly higher than that of the ANN model in summer when EPs touch their peak values. Analysis of prediction errors and outliers also confirms the higher accuracy and lower uncertainty of CNN-LSTM model compared to ANN. Consequently, from the viewpoint of hydropower producers, the CNN-LSTM model is much more reliable, especially during periods of peak energy consumption with high energy prices in summer which is a challenging time to forecast MDEP.