This paper presents a novel approach to predicting the pressure and flow rate of flexible electrohydrodynamic (EHD) pumps using Kolmogorov-Arnold Networks (KAN). Inspired by the Kolmogorov-Arnold representation theorem, KAN replaces fixed activation functions with learnable spline-based activation functions, enabling it to approximate complex nonlinear functions more effectively than traditional models like Multi-Layer Perceptron (MLP) and Random Forest (RF). The study evaluated KAN on a dataset of flexible EHD pump parameters and compared its performance against RF and MLP models. KAN achieved superior predictive accuracy, with Mean Squared Errors (MSE) of 12.186 and 0.012 for pressure and flow rate predictions, respectively. The symbolic formulas extracted from KAN provided insights into the nonlinear relationships between input parameters and pump performance. These findings demonstrate that KAN offers exceptional accuracy and interpretability, making it a promising alternative for predictive modeling in electrohydrodynamic pumping. The study introduced a novel application of KAN for predicting the performance of flexible EHD pumps and providing insights into the relations of input parameters in determining the output performance of flexible EHD pumps. The KAN model was trained on a dataset of 98 samples, each containing five input features and two output features. The input features include channel height, electrode overlap, voltage, electrode gap, and apex angle, while the output features are the maximum pressure and maximum flow rate of the flexible EHD pumps. The KAN model achieved remarkable predictive accuracy, with MSE values of 12.186 and 0.012 for pressure and flow rate predictions, respectively. These results significantly outperform RF and MLP models. The model's ability to accurately predict pressure with a low MSE makes it particularly valuable for applications in soft robotics and biomedical devices, where precise pressure control is essential. The KAN model also provides interpretable symbolic formulas that reveal the mathematical relationships between input variables and output metrics, making it particularly valuable for applications requiring a deep understanding of these underlying relationships.This paper presents a novel approach to predicting the pressure and flow rate of flexible electrohydrodynamic (EHD) pumps using Kolmogorov-Arnold Networks (KAN). Inspired by the Kolmogorov-Arnold representation theorem, KAN replaces fixed activation functions with learnable spline-based activation functions, enabling it to approximate complex nonlinear functions more effectively than traditional models like Multi-Layer Perceptron (MLP) and Random Forest (RF). The study evaluated KAN on a dataset of flexible EHD pump parameters and compared its performance against RF and MLP models. KAN achieved superior predictive accuracy, with Mean Squared Errors (MSE) of 12.186 and 0.012 for pressure and flow rate predictions, respectively. The symbolic formulas extracted from KAN provided insights into the nonlinear relationships between input parameters and pump performance. These findings demonstrate that KAN offers exceptional accuracy and interpretability, making it a promising alternative for predictive modeling in electrohydrodynamic pumping. The study introduced a novel application of KAN for predicting the performance of flexible EHD pumps and providing insights into the relations of input parameters in determining the output performance of flexible EHD pumps. The KAN model was trained on a dataset of 98 samples, each containing five input features and two output features. The input features include channel height, electrode overlap, voltage, electrode gap, and apex angle, while the output features are the maximum pressure and maximum flow rate of the flexible EHD pumps. The KAN model achieved remarkable predictive accuracy, with MSE values of 12.186 and 0.012 for pressure and flow rate predictions, respectively. These results significantly outperform RF and MLP models. The model's ability to accurately predict pressure with a low MSE makes it particularly valuable for applications in soft robotics and biomedical devices, where precise pressure control is essential. The KAN model also provides interpretable symbolic formulas that reveal the mathematical relationships between input variables and output metrics, making it particularly valuable for applications requiring a deep understanding of these underlying relationships.