This research introduces an enhanced LSTM framework utilizing PSO for forecasting the opening value of the NSE index. This study introduces the LSTM and augmented PSO-LSTM framework for stock price prediction, utilizing the notion of time series. The enhanced PSO-LSTM technique utilizing PSO to enhance the weights of Long Short Term Memory framework, thereby enhancing prediction accuracy. PSO is employed to adjust weights of the LSTM framework, thereby diminishing forecasting error. Following the preliminary processing of past stock data, which encompasses starting price, closing cost, highest cost, lowest rate, as well as regular volume, we develop the LSTM utilizing time series derived from this past dataset. Ultimately, we implement the suggested LSTM to forecast opening value of NSE index. The PSO optimization algorithm, applied to the LSTM model through empirical study, efficiently identifies the best neural network weights, minimizes the loss function, facilitates speedy fitting, and yields more accurate predictions.