Cleft Lip/ Palate is the most common facial birth defect, which is fully reversible through surgeries and rehabilitation starting at an average age of 3 months. Cleft surgeries aim to enhance facial symmetry and nasolabial appearance, but currently lack standardized procedures, often relying on the surgeon’s intuition and experience. Cleft surgical success is assessed using subjective scores based on aesthetics and facial symmetry, which can be biased and inconsistent due to the absence of objective measurements. In this study, we propose a methodology to develop objective Cleft Surgical Success Scores. For this purpose, we employ GAN-Inpainting models to generate a normalised post-operative lip area, serving as a reference image. An Image Similarity Score is proposed to evaluate differences between the post-operative and synthesized reference images, providing an objective metric that penalises any post-surgical deformities or scars. This approach is designed to standardize the evaluation of surgical outcomes, minimizing human bias and enhancing the reliability of surgical success assessments.