Brain tumors are among the most fatal and devastating diseases, often resulting in significantly reduced life expectancy. An accurate diagnosis of brain tumors is crucial to devise treatment plans that can extend the lives of affected individuals. Manually identifying and analyzing large volumes of MRI data is both challenging and time-consuming. Consequently, there is a pressing need for a reliable deep learning (DL) model to accurately diagnose brain tumors. In this study, we propose a novel DL approach based on transfer learning to effectively classify brain tumors. Our novel method incorporates extensive pre-processing, transfer learning architecture reconstruction, and fine-tuning. We employ several transfer learning algorithms, including Xception, ResNet50V2, InceptionResNetV2, and DenseNet201. Our experiments used the Figshare MRI brain tumor dataset, comprising 3,064 images, and achieved accuracy scores of 99.40%, 99.68%, 99.36%, and 98.72% for Xception, ResNet50V2, InceptionResNetV2, and DenseNet201, respectively. Our findings reveal that ResNet50V2 achieves the highest accuracy rate of 99.68% on the Figshare MRI brain tumor dataset, outperforming existing models. Therefore, our proposed model's ability to accurately classify brain tumors in a short timeframe can aid neurologists and clinicians in making prompt and precise diagnostic decisions for brain tumor patients.