Gastrointestinal stromal tumor (GIST) presents significant challenges in its precise diagnosis and prognostic evaluation. In recent years, radiomics has enabled the construction of various high-performance predictive models by extracting and analyzing high-dimensional quantitative features from medical images such as endoscopic ultrasound, computed tomography, and magnetic resonance imaging, in combination with machine learning and deep learning algorithms. These models have demonstrated considerable values in the identification and differential diagnosis of GIST, risk stratification, proliferation activity prediction, inference of gene mutation types, assessment of therapeutic response, and prediction of recurrence and metastasis risks. This article provides a systematic review of the advances in radiomics-based predictive models for the diagnosis and prognosis of GIST.