Gastric cancer is a common and highly lethal cancer in China, where early diagnosis and treatment are crucial for patient prognosis. Currently, clinical physicians mainly rely on endoscopy, imaging, and histopathology for the diagnosis of gastric cancer, which are often image-based methods. These methods, however, are highly subjective, time-consuming with low accuracy. In recent years, artificial intelligence (AI) technology has gradually become more prevalent in the medical field. AI systems based on deep learning can extract features from large volumes of medical image data, assisting physicians in making rapid and accurate diagnoses. This article reviews the research progress of image-based AI in gastric cancer and discusses its future development directions.