Deep learning provides a wide array of applications, particularly in the fields of health and diagnostics. The purpose here is to accurately predict the existence of pneumonia based on an X-ray scan. This is particularly advantageous in the present circumstances, as COVID-19 is recognized for its propensity to induce pneumonia. Pneumonia is a viral infection that commonly affects many people, particularly in underdeveloped and impoverished nations where there is widespread pollution, overcrowding, filthy living circumstances, and a lack of healthcare infrastructure. Pneumonia leads to the development of pericardial effusion, a condition characterized by the accumulation of fluids in the chest, resulting in difficulties with inhalation. Detecting the diagnosis of pneumonia promptly is a challenging task, crucial for accessing treatment resources, and enhancing the likelihood of survival. Multiple methods can be used to detect pneumonia, including CT scans, pulse oximetry, and others. However, the most frequently employed method is X-ray tomography. However, the technique of analyzing chest X-rays (CXR) is challenging and prone to subjective variability. This study employs six deep learning models such as CNN, Densenet, VGG16, ResNet50, InceptionNet, and MobileNet models to identify and categorize pneumonia from CXR. The MobileNet achieves superior performance, with a 96.21% accuracy score.