The main goal of this work is to create and assess a deep learning model for classifying cervix types from colposcopy pictures. 200 photos total, separated into three groups (Type 1, Type 2, and Type 3). and identified by medical professionals make up the dataset. Using the pre-trained weights from ImageNet, we used three cutting-edge deep learning architectures: VGG19, Inception v3, and ResNet50. The models were put to the test, trained, and verified using common assessment metrics including F1- score, accuracy, precision, and recall. The findings show encouraging performance, with an accuracy of 83.5% attained overall for all cervix types. The model's sensitivity varied from 75.8% to 87.0%, suggesting its efficacy in identifying positives for each cervix type, while its specificity ranged from 84.7% to 93.0%, displaying its capacity to properly identify negatives. By advancing automated cervix- type categorization systems, this study may help diagnose cervical cancer earlier and enhance clinical decision-making procedures.