This investigation into breast cancer diagnosis from mammography pictures uses deep learning algorithms MobileNetV2 and VGG. This research used a publicly accessible dataset of mammograms to train classification algorithms to distinguish between cancerous and noncancerous tissue. Both algorithms' efficacy was measured by a variety of criteria, including accuracy, precision, recall, and F1 score. High levels of accuracy in breast cancer diagnosis were reached by both MobileNetV2 and VGG, with VGG marginally beating MobileNetV2. This work adds to the growing body of evidence that deep learning algorithms may be useful for enhancing the accuracy and efficiency of breast cancer diagnosis using mammography pictures.