We find that the additivity of quantum information channels enables one to introduce a quantum classifier or a quantum decision maker. Proper measurement and sensing of temperature are of central importance to the realization of nanoscale quantum devices. Minimal classifiers may constitute the basic units for the physical quantum neural networks. We introduce a binary temperature classifier quantum model that operates in a thermal environment. In the present study, first the mathematical model was introduced through a two-level quantum system weakly coupled to the thermal reservoirs and it was demonstrated that the model faithfully classifies the temperature information of the reservoirs in the thermal steady state limit. A physical model by superconducting circuits composed of transmon qubits was also suggested.