<p>The melting point of primary slag is a key factor in blast furnace operation, influencing fluidity, bonding zone thickness, differential pressure, and permeability. However, few models predict the primary slag melting point, and quantifying the effect of charge structure is challenging. In this study, ten plants were selected for ore preparation, the composition of primary slag was calculated, and the primary slag melting point was calculated by thermodynamic calculation software. The characteristic variables of primary slag melting point were correlated and analyzed. The primary slag melting point was mainly affected by the mass fraction of MgO and showed a significant positive correlation trend. In addition, a primary slag melting point prediction model was developed based on two algorithms: Gaussian process regression (GPR) and random forest regression (RFR). Both algorithms showed high accuracy in primary slag melting point prediction, and the accuracy of their prediction results within the ± 1&#xa0;°C error range exceeded 95%. However, the mean absolute error (MAE) and root-mean-square error (RMSE) of the Gaussian process regression algorithm were 0.26 and 0.49, respectively, which were lower than those of the random forest regression algorithm, suggesting that it is more advantageous in prediction performance.</p> Graphical Abstract <p></p>