Exploiting the energy flexibility resulting from thermal loads management in buildings is one of the most promising solutions to contribute to the energy transition targets. To offer energy flexibility to the grid, the building should meet the requirements: (i) it needs electrically powered generation (e.g., through Heat Pumps) and (ii) advanced control techniques (e.g., Model Predictive Controls) must be implemented. In recent years, the scientific community has produced many studies demonstrating the potential of Model Predictive Controls combined with Heat Pumps to exploit energy flexibility in buildings. However, a large-scale deployment of such control techniques is still far off, both because of the not yet widespread use of Heat Pumps and the computational challenges involved in implementing them. The aim of this study is to contribute to the deployment of advanced control techniques for Heat Pumps systems in buildings by simplifying their implementation. At this aim, validated archetypes of Model Predictive Controls for Heat Pumps in residential buildings are proposed. The availability of archetypes can greatly facilitate the practical application of Model Predictive Control. In fact, they are adaptable to the characteristics of different buildings and Heat Pump installations and their structure is designed to have an acceptable trade-off between calculation time and prediction reliability. The archetypes cover three different types of heating systems: low temperature radiators without (i) and with (ii) integrated heat storage devices and (iii) underfloor heating system. All archetypes are applied to a real Heat Pump and validated through experimental campaigns. During the experimental test all the archetypes proved to be effective in controlling the real system and the results showed good reliability of the prediction model in the control (for all archetypes a Root Mean Square Error lower than 0.44 °C was obtained) and an optimizer success rate in Model Predictive Controls greater than 91 %. The archetypes proposed are provided as open-source tools that can be reused for similar cases to facilitate Model Predictive Controls implementation in heat pump systems.