The objective of this research is to study the heat insulation performance of ceramic additives paint (CAP) with the variation of electric power. The thickness of the closed media was 13 cm and the internal dimensions were 1m × 1m × 1m. The two closed media were painted by the regular paint and the ceramic content of insulation layer was 18%. The temperature of internal closed media was controlled by thermostats to be in two levels, namely 28 °C and 40 °C. The results indicated that a considerable electric energy gain can be achieved by CAP in comparison with the regular paint. Additionally, it was observed that in indoor closed media with ceramic additive paint, the internal temperature reached 40 °C and 28 °C in a sufficient short period of time. In this paper, adaptive neuro-fuzzy inference system (ANFIS) is used to predict the energy consumption and closed media temperature. The ANFIS model has been applied for the training of the fuzzy system and the test set was used to evaluate the performance of the system using statistical parameters, which were R2 > 0.95, MAE < 0.96, MSE < 1.53, RMSE < 1.28, and MAPE < 0.98. The results approved that the predicted values from the model were in a good agreement with the experimental data.