The current dangerous viral disease is COVID-19. This viral disease can spread quickly and not only to humans, but animals can also contract the COVID-19 virus disease. Until now, the spread of the COVID-19 virus has not been able to stop. The COVID-19 virus disease is mostly caused by lung infections. To diagnose COVID-19 virus disease with more effective imaging using CT scan images. DenseNet121, MobileNet, Xception, InceptionV3, ResNet50V2, and VGG19 models to check their accuracy in image recognition. To analyze the model performance, 1361 samples from CT scans were collected from the Kaggle warehouse. The DenseNet121-MobileNet model has a sensitivity of 99.76%, a specificity of 99.63%, and an accuracy of 99.76% with a computation time faster. This work focuses only on the methods used to detect patients with COVID-19 in the lungs but does not mention any medical accuracy.