The binding affinity of a protein-ligand denotes the degree of strength or affinity between them and the specificity of the interaction between a protein and a ligand, where a ligand is typically a smaller molecule that binds to a specific site on the protein. This binding affinity is a critical aspect of various biological processes, including enzymatic reactions, cellular signaling, and the development of therapeutic drugs. The term describes how tightly the ligand is bound to the protein and is often quantified using measures such as the dissociation constant (Kd). A lower Kd indicates a higher binding affinity, suggesting a stronger and more stable interaction. Assessing binding affinity is vital in predicting how drugs behave in the body, influencing aspects such as absorption, distribution, metabolism, and elimination. This information is essential for designing drugs with optimal therapeutic profiles and predicting their effects on the body. The application of deep learning in predicting protein-ligand binding affinity represents a cutting-edge approach with the potential to significantly impact drug discovery and computational biology. Here convolution neural network (CNN) model was developed to predict protein-ligand binding affinity. This innovative approach involved using encoder and decoder components to handle the unique representations of proteins and ligands. The proteins were encoded in a fast format, while the ligands were represented in a smile format. The application of a convolution neural network (CNN) model to predict protein-ligand binding affinity yielded promising results, as evidenced by the performance metrics obtained. The mean absolute error (MAE) of 0.3534, mean 374 square error (MAE) of 0.2515, root mean squared error (RMSE) of 0.5014, and R-squared of 0.4030.