This study presents a software tool Grav3CH_inv developed to estimate the three-dimensional depth structure of sedimentary basins from their gravity data through an iterative process. The algorithm linked to the developed code operates recursively both in the wavenumber domain and in the space domain based on a triple method combination. The modelling strategy allows considerations of the model space with an exponential increase in density with depth and also with density variations in the horizontal direction. The accuracy of computation of gravity anomalies in the FFT-based forward procedure is also increased by using the shift-sampling technique which minimizes discretization effects during transformations between the space and wavenumber domains. Given the observed gravity anomalies and the density design of the basin, the iterative procedure performs automatically until the goodness of the fit between the observed and modeled anomaly, either the root-meansquare error or the largest error is below its pre-assigned value. As an advantage, the computing time is acceptably short for such a kind of an modelling problem. The GUI-enabled interactive control functions of the Grav3CH_inv code allow users to set optional settings, style of outputs and export formats, and facilitate operations without requiring coding expertise to perform the relevant procedures of the algorithm. The feasibility and accuracy of the proposed software is demonstrated by evaluating the synthetically produced gravity anomalies of various 3D basin models and also by analyzing an actual gravity data from the Los Angeles basin, California.