The increasing availability of high-fidelity aerodynamic databases has fostered the development of surrogate models capable of delivering accurate predictions at a significantly reduced computational cost. In this context, this work presents the development and evaluation of a Support Vector Machine (SVM) surrogate model for the aerodynamic prediction of the NASA/Boeing Common Research Model wing–body–pylon–nacelle (CRM WBPN) configuration. The study is part of the regression challenge organized by ONERA, which provides a common, publicly available database and benchmarking environment for machine-learning-based aerodynamic modeling. The surrogate model is trained using the CRM WBPN database prepared by ONERA which contains high-fidelity wall pressure distribution data associated with a range of flight and geometric parameters. Unlike most of the competing approaches in the challenge, which rely predominantly on Multilayer Perceptron (MLP) neural networks, this work focuses exclusively on an SVM regression model implemented in Python. Particular attention is given to the selection of hyperparameters and the generalization capabilities of the model, with the aim of accurately extracting and predicting the aerodynamic characteristics of interest from the provided input variables. The performance of the proposed SVM surrogate is evaluated and analyzed within the context of the ONERA 468 CRM Wall Distribution Regression Challenge hosted on Codabench. Although the results are not directly comparable to those of the top-ranking neural-network-based models, the SVM demonstrates competitive predictive accuracy and robust behavior in the test cases. These findings highlight the relevance of SVMs as viable and efficient alternatives to deep learning approaches for aerodynamic surrogate modeling, especially in scenarios with limited training data or where model interpretability and stability are of interest.