To predict stock prices accurately, historical data plays a crucial role. However, if stock returns are considered random, then predicting tomorrow’s market price would be simply based on today’s price plus a negligible increase or decrease. In time series analysis, extracting meaningful features is important to comprehend patterns and trends over time. There exist various techniques for feature extraction that aid in identifying significant characteristics of the data and enhancing the precision of predictions. In the paper, we introduce a pair of models for deep learning, the Gated Recurrent Unit (GRU) and Convolutional Gated Recurrent Unit (ConvGRU), for stock price prediction. Traditional analysis methods may not be sufficient for uncovering hidden information that contributes to stock price fluctuations. Therefore, we propose the use of these models to enhance our ability to predict stock prices. To test the hypothesis, two deep learning models were trained using historical stock price data that underwent preprocessing: GRU and ConvGRU. The GRU model was developed to predict future prices based solely on temporal information, while the ConvGRU model was trained to process both temporal and spatial information in the input data. This methodology was employed to determine whether incorporating external information could enhance stock price prediction accuracy. According to the experimental outcomes, the sheer performance of the proposed GRU model surpassed that of conventional machine learning models such as linear regression and ARIMA. Hence, we can conclude that the GRU model exhibits the potential to become a robust tool for predicting stock prices and can be utilized in various financial and investment-related domains. The absolute quality of the input data stack, the point choice of hyperparameters, and the length of the predicted time frame are some of the variables that might affect how well the GRU model predicts stock values. Hence, in general, The GRU model can be an effective tool for forecasting stock values, but investment decisions shouldn’t be made only on the basis of it. There are other factors also that affects the stock market, such as economic indicators, geopolitical events, and market sentiment. So, the prediction accuracy is greatly reduced hence use of GRU and LSTM models was necessary for effective prediction.