This research aims to benchmark and evaluate the performance of various foundation models in time series forecasting especially in the domain of demand forecasting. Study has used real life demand data set from a e-grocery innovator. This study uses already established statistical, machine learning and deep learning algorithms and compares their forecasting performance with some of the popular pretrained models for time series forecasting and evaluates multiple accuracy metrics to establish a credible framework for comparison and benchmarking. This study has shown that machine learning models are usually the best performing models across various error metrics and various time granularities. However, it was also noticed that pretrained models occasionally perform better than all other established models. Code and data used in the study is available at https://anonymous.4open.science/r/CriticalEvaluation-ofFoundationalModelsinDemandForecasting-BB71/