Alzheimer’s disease is a devastating neurological disorder that affects millions of people worldwide. It is characterized by the progressive death of brain cells, leading to brain shrinkage and a decline in cognitive abilities, behavior, and social skills. Therefore, early diagnosis is crucial for optimizing care, improving quality of life, and advancing our understanding of the disease. This study presents a model that utilizes transfer learning and fine-tuning techniques to distinguish between normal and Alzheimer-affected brain images. The dataset includes 6336 MRI images, categorized into three classes: non-demented, mild-demented, and very mild-demented. The results demonstrate that using the VGG16 model’s transfer learning and fine-tuning techniques achieves promising outcomes, with accuracy and F1-score of 86%, 92%, 94%, and 98%, respectively. The proposed model performs better than some state-of-the-art methods, highlighting its potential to aid in early Alzheimer’s diagnosis. Furthermore, this approach’s ability to differentiate between various stages of Alzheimer’s disease can improve patient outcomes, allowing for earlier interventions and better planning for patient care. Overall, the proposed model’s performance demonstrates its potential as a valuable tool in diagnosing and treating Alzheimer’s disease.