In the work herein has been carried out the performance prediction of the solar heat supply system with the thermal pump using artificial neural networks. Techniques of the artificial intellect (AI) become useful as an alternative approach to the conventional methods or as components of integrated systems for solving the complex practical problems in different fields and become more and more popular now-a-days. Experimental works on manufacturing the thermal pumps have been fulfilled for the depth of 0.5m, with performance parameters–heating time, pump efficiency and work discharge. By means of artificial neural network (ANN) the model has been developed for assessing the performance of the solar heat supply system with the thermal pump. The ANN model is quite well trained, it accelerates the training process, and further it has been said that the ANN is quite well trained for the solar heat supply systems.