Cavities are among the most common dental health problems and significantly impact the quality of life, particularly in developing countries. Early detection of dental caries is a crucial step in preventing further complications; however, conventional methods such as clinical examinations and radiography are often inaccessible due to infrastructure and cost limitations. This study aims to develop an Artificial Intelligence (AI) model based on MobileNetV3 Small for detecting dental caries using images captured with a basic smartphone camera. MobileNetV3 Small was selected for its high computational efficiency and ability to operate on low-specification devices. The dataset used comprises 1,200 dental images, including both healthy teeth and teeth with cavities. The images were taken under varying lighting conditions and resolutions to reflect real-world scenarios. The model was trained using transfer learning and evaluated on a validation dataset using accuracy, sensitivity, and specificity metrics. The results demonstrated that the model achieved 90% accuracy, 90% precision, and 90% recall, highlighting its potential for real-time applications. These findings suggest that MobileNetV3 Small can serve as a practical, cost-effective, and accessible solution for early detection of dental caries using everyday devices like smartphones. This technology has the potential to improve access to dental health services, support early detection initiatives, and reduce the prevalence of dental caries. This research provides a foundation for further development of AI applications in healthcare, particularly in developing countries.