The COVID-19 virus has spread globally, wreaking havoc on people's lives, economies, and healthcare systems.The best strategy to stop the spread of COVID-19 is by rapid and accurate diagnosis, hence many different methods have been used. Pulmonary collapse and lung damage are potential complications of COVID-19. Consequently, chest X-ray has evolved into one of the dependable diagnostic technologies coupled with AI methods that may be used to verify physicians' assessments. Our study suggested two models for COVID-19 detection from chest X-ray images: one based on deep learning's Convolutional Neural Network (CNN), and the other on transfer learning's InceptionV3. This study draws from many datasets totaling 1553 Chest X-ray pictures. Our suggested CNN architecture, based on deep learning, has the greatest training accuracy (79.74%) and validation accuracy (84.92%). In contrast, InceptionV3's transfer learning-based architecture had the greatest training and validation accuracy at 85.41% and 85.94%, respectively.