The rapid advancement of digital financial technologies has accelerated the adoption of cryptocurrencies, with Bitcoin emerging as the dominant asset characterized by extreme price volatility and investment risk. Despite extensive studies on Bitcoin forecasting, existing predictive models remain limited in capturing long-term volatility dynamics and complex temporal dependencies, leading to unstable performance under fluctuating market conditions. This study addresses this gap by developing a deep learning-based forecasting framework using the Long Short-Term Memory (LSTM) algorithm integrated with a real-time web-based application. Historical Bitcoin price data were preprocessed through Min–Max normalization and transformed into time-series sequences using sliding window techniques. The proposed model consists of two stacked LSTM layers with 100 hidden units each, followed by a dense output layer, and was trained using the Adam optimizer with early stopping to prevent overfitting. Model performance was evaluated using Root Mean Squared Error (RMSE), Mean Squared Error (MSE), and Mean Absolute Error (MAE). The experimental results demonstrate that the proposed LSTM model achieved a Test MAE of 2.27%, indicating substantially higher accuracy compared to conventional statistical forecasting approaches reported in prior studies. The model effectively tracks long-term price trends, although extreme short-term spikes remain challenging due to inherent market volatility. Furthermore, the integration of the trained model into a Flask-based web application enables interactive real-time price prediction, representing a practical innovation beyond offline forecasting models. Overall, this research demonstrates the effectiveness of deep learning for supporting cryptocurrency investment decisions in real-world practice.