Stochastic volatility analysis is a sophisticated method for forecasting gold prices, accounting for the fact that volatility which is the degree of variation in gold prices is not constant over time but rather fluctuates unpredictably. This approach can be particularly effective because gold prices are often subject to sudden and unpredictable changes due to various economic, political, and market factors. Different statistical and machine learning models are available to predict the Gold price such as ARIMA, GARCH, SVM, Random Forest, GRU and CNN. In this paper LSTM model is implemented to forecast the Gold Price on the basis of 10 years of data from January 2013 to July 2023. Ater the implementation the value of Mean Absolute Error obtained is 1775.6294. In the future, this model can be extended to different versions of LSTM.