<p>In recent decades, a vast network of Continuously Operating GNSS Reference Stations (CORS) has been established worldwide. Initially designed to enhance the accuracy and reliability of global navigation, these networks are increasingly used for meteorological applications. The precise station coordinates, accurate to within a few millimeters, enable the estimation of atmospheric effects on the radio signal delays. This approach supports meteorological studies and also demonstrates the versatility of GNSS-based methods in various geoscientific applications. Zenith tropospheric delays (ZTD) estimated from GNSS observations are often assimilated in numerical weather models since they are linked to the vertical integral of atmospheric water vapor (IWV). However, GNSS observations stemming from a dense network can be used to model the spatial distribution of atmospheric water vapor (AWV) using tomographic reconstruction. The toolbox is designed to reconstruct the 3D atmospheric wet refractivity field based on the estimated zenith tropospheric delays and the tropospheric gradients. The 3D wet refractivity model can be converted to atmospheric water vapor (AWV) density when the temperature profile is available. The paper briefly overviews the principles of tomographic reconstruction using a multiplicative algebraic reconstruction technique (MART) and introduces an open-source toolbox written in Python that performs the full data processing chain, including data quality checks. The toolbox can be integrated into near real-time GNSS data processing environments. To verify the accuracy of the reconstructed wet refractivity model, a quality control method is applied using radiosonde (RS) measurements. The results show that the standard deviation of the reconstructed 3D refractivity model is 10 ppm (or N-units) below 3&#xa0;km altitude and 0.3&#xa0;ppm at an altitude of 10&#xa0;km.</p>