Automatic modulation classifiers (AMC) have emerged as indispensable tools to tackle these communication challenge effectively. While numerous researchers have made significant strides in modulation classification using various deep learning and machine learning (ML) techniques. Most existing AMC models have primarily focused on single-carrier modulation signals. Given the prevalence of multicarrier modulation (MCM) in the current wireless communication infrastructure, the necessity for MCM classification methods has become paramount. This paper introduces a robust model for classifying MCM signals, incorporating a novel approach that integrates Short-Time Fourier Transform (STFT) spectrograms and the Xception model. The proposed model encompasses the classification of six distinct types of MCM signals. Experimental assessments demonstrate the efficacy of the proposed approach, achieving a superior classification accuracy of 89% at – 8 dB and 97% at 8 dB SNR.