In women, cervical cancer is a somewhat prevalent illness. Lives can be saved by early discovery and precise diagnosis. Traditional methods of cervical cancer detection involve manual screening of cell images and manual classification. This task leads to a delay in diagnosis and might also lead to misdiagnosis. Therefore, a computer-aided diagnosis technique that can perform disease diagnosis efficiently and precisely is needed. Cervical cancer can be detected by a cell image, a colposcopy, or a Pap smear test. Pap smear tests are more popular because they are less expensive and provide a painless diagnosis. Size, shape, and color can all be used to determine the degree of anomalies in the cells on Pap smear cluster slides. The classification and segmentation of Pap smear images can be accomplished using various deep-learning methods. In our research paper, EfficentNet, Resnet-50, Vgg-19, Densenet-121, and MobileNet are the five deep-learning models that are applied to the SIPAKMED dataset. An ensemble of deep-learning models was used to enhance the overall performance of the dataset. The model created using all five deep-learning models achieves an accuracy of 98% on the dataset.