Much attention has recently been paid to the exponential rise of Bitcoin and other cryptocurrencies. Bitcoin is the most widely used cryptocurrency in the world. An increasing body of research examines cryptocurrencies from two distinct perspectives, spurred by the market's heightened correlation and volatility as well as the notable rise in market capitalization that has given rise to contemporary financial products like futures and options. The term spillover effect describes the phenomenon where changes in one cryptocurrency's performance or value affect other cryptocurrencies. In this work, the spillover effect in cryptocurrencies is analysed using four distinct deep learning models: gated recurrent unit (GRU), long short-term memory (LSTM), convolutional neural network (CNN), and bidirectional LSTM (Bi-LSTM). The models are compared with the root mean squared error (RMSE) and mean absolute error (MAE) scores. In this work, the BTC-USD dataset is taken from Kaggle competitions from January 2012 to September 2020. The RMSE score of the Bi-LSTM model is 6.538, and the MAE score is 1.471. As compared to other models, the Bi-LSTM model performs well. The spillover effects can improve market efficiency by swiftly spreading knowledge among different cryptocurrency platforms.