High entropy alloys constitute a field of great interest in Materials Science due to their extraordinary mechanical properties. There are multiple combinations for the systhesis of new kind of these materials, so the implementation of Artificial Intelligence provides a great tool for accelerating the initial phases of HEAS’s design. As Neural Network a CTGAN was used, which is a GAN that can be fed with a tabular dataset and some of that data introduced as a condition. GANs consist of two neural networks, the generator that generates synthetic compounds and the discriminator that learns from a specific data structure containing real samples, so then it is able to classify the generated ones into real or fake. After the training process, CTGAN’s output is a set of generated compounds with the same features associated as for the real ones. Finally, validation was done with DFT databases.