Seismic surveying represents an efficient and pivotal tool in the exploration of hydrocarbon reserves, consistently drawing the focus of the upstream oil and gas industries. Seismic inversion, one of the most critical methods for delving into the characteristics of subsurface geological layers, poses a formidable challenge. In this study, we’ve departed from the conventional approach of merely enhancing or combining existing seismic inversion methods. Instead, we’ve employed a novel generative-adversarial network (GAN) algorithm, which is a deep learning algorithm, meticulously trained for the seismic inversion process. This innovative approach aims to omit several pressing challenges, including the computation of the inversion matrix, initialization of the wavelet, and navigating the constraints of the limited frequency band of seismic amplitudes in seismic inversion. This study has yielded remarkable results. Through the application of the generative-adversarial deep learning algorithm, we’ve not only conquered the aforementioned challenges but have also achieved exceptional quality and precision in our results. We conducted seismic inversion using real data from an oil field, achieving an impressive accuracy rate of 97.5%. This accuracy percentage is validated based on both the validation data and the mean squared error (MSE), reinforcing the robustness of the proposed approach. Furthermore, the acoustic impedance of the five test wells consistently measured below 0.125 units, highlighting the excellence of our outcomes. The correlation coefficient among these test wells ranged from a minimum of 96% to a maximum of 99%. In contrast, the acoustic impedances obtained through the band-limited method and the model base method displayed correlation coefficients of 71% and 83%, respectively. The utilization of the generative-adversarial algorithm in the inversion process underscores its contemporary efficiency. It holds the potential to entirely modernize and refine the conventional seismic inversion process, ushering in a new era of seismic exploration and inversion.