The classification of brain tumors is a vital aspect of disease analysis and monitoring. Magnetic resonance imaging (MRI) is the most commonly employed method for detecting brain tumors. However, it is noted that the interpretation of MRI scans can be subject to variability due to different evaluation formulations used by specialists, which can lead to erroneous results. Furthermore, the identification of a tumor's presence alone is insufficient. Prompt treatment initiation is equally critical, and thus, determining the tumor's type is of paramount importance. In this study, we present a machine-learning classifier for brain tumor classification that identifies the tumor as benign or malignant. To achieve faster, more reliable, and unbiased tumor detection, we evaluated several deep learning architectures using a dataset of 187,092 images for training and 45,183 images for testing.