Natural gas is paving the way in experiencing a high-tech renaissance for the world, by benefiting consumers, creating jobs, and revolutionising the energy landscape of the world. The need to effectively estimate natural gas prices has become critical for stakeholders involved in energy planning, commodities trading, and investment choices on natural gas markets, given their enormous economic impact. By incorporating traditional and hybrid models, natural gas prices are predicted. This study specifically examines deep learning models and compares them to determine which strategy works better. N-BEATS, a unique deep learning model, is applied to forecast natural gas prices and deep learning models such as the dense model, LSTM-powered RNN, and Naïve model, are trained on historical prices of natural gas in the US from year from January 10, 1997, to January 5, 2024 and employed to predict prices in future. N-BEATS leads in performance measures, surpassing all the other models.