We explain how thermodynamic cost can use for diagnosing optimal dense coding. We present a quantum channel where is included two initially uncorrelated thermal quantum systems to reveal the optimal dense coding using thermodynamic cost. The interest in dense coding brings into quantum correlation calculation. At first, the quantum Fisher information and spin squeezing are used to quantify the correlation dynamics over the system. The system reveals that the thermal evolution of quantum correlations depends crucially on specific energy and temperature. Also, they can be utilized as control parameters for optimal dense coding. Several interesting features of the variations of the energy cost and the dense coding capacity are obtained. It can keep its valid capacity value in a broad range of temperatures by increasing the energy value of excited states. Also, we can identify valid dense coding with the help of calculating the energy cost of the system. Using this approach, identifying a critical point of this model in dense coding capacity quality can be very effective.