Helical-coiled once-through steam generators (H–OTSG) are distinguished by their compact configuration, high heat transfer efficiency, and reliability, which contribute to their extensive use in advanced small reactors. Nonetheless, the intricate three-dimensional structure and the complex secondary flow dynamics present significant challenges in accurately predicting the heat transfer coefficient (HTC) of helically coiled tubes. This study is based on four helically coiled tubes from SJTU-NETH covering all heat transfer zones. Spearman correlation was used to analyze the input parameters. A predictive model was developed based on the selected features using artificial neural networks (ANNs). The resulting model demonstrates strong predictive capabilities across the complete parameter range, and its effectiveness in accurately capturing the physical trends of HTC variations during the flow process within the coil has been validated. This research enhances the understanding of heat transfer dynamics in helical-coiled tubes and provides valuable insights for optimizing the design of H–OTSG. Additionally, the developed model improves the predictability of the heat transfer coefficient, thereby contributing to the advancement of safer and more efficient thermal systems in advanced small reactors.