With the growing popularity of electric vehicles, as a means of addressing climate change, there are concerns about their adoption affecting electric grid management. Consequently, predicting charging demand is a timely and valuable research effort. There has been plenty of research on the forecasting of energy load in transportation. However, there are limited efforts to compare multiple methods over different temporal and spatial horizons, across a variety of cities. This thesis investigates the effectiveness of five time series forecasting models, ranging from traditional statistical ones to more recent machine learning and deep learning methods. The predictions cover short-, mid- and long-term forecasting scenarios, examining performance from individual charging station level through regional aggregations to city scale implementations. All models were implemented in lightweight form, with no hyperparameter tuning, limited epochs and simple, pre-defined architectures. According to the results, ARIMA outperforms the other, more complex models across all configurations. Despite its simplicity, it achieves the lowest prediction errors, highlighting its robustness and timelessness.