Accurate and timely brain tumor diagnosis is critical for effective treatment and improved patient outcomes. This study investigates the application of deep learning to automate the classification of four common brain tumor types (glioma, meningioma, notumor, and pituitary) using MRI images. We propose a novel ensemble model, combining two meticulously fine-tuned EfficientNetB3 and MobileNet architectures, to enhance classification accuracy. The model is trained on a dataset of 7023 MRI images, incorporating a comprehensive data augmentation strategy to address data scarcity challenges. Our proposed approach achieves a remarkable accuracy of 98.86%, surpassing the performance of individual CNN models and recent studies in this field. Our findings demonstrate that ensemble learning and data augmentation significantly improve model performance and generalization. This research contributes to the growing evidence supporting AI’s transformative potential in healthcare, particularly for improving the accuracy and efficiency of brain tumor diagnosis.