This paper presents a hierarchical multiscale approach based on a micro-kinetic model enabling to predict temperature, pressure and species generated during the thermite reaction of Al nanoparticles mixed with CuO nanoparticles. Overall, our phenomenological model integrates and combines series of complex atomistic mechanisms, e.g. diffusion and phase transformation, gas phase reactions and interphase exchange mechanisms, in particular molecular condensation, evaporation and decomposition. Thermodynamics considerations as well as Density Functional Theory (DFT) calculations are used to implement rate equations expressing the complex reactions at solid/liquid/gas interphases. We demonstrate that the model can predict the pressure-time dependence, the different phases and compositions with good accuracy at significantly low computational cost. The influence of AI and CuO particle size, compaction or density, alumina shell thickness and stoichiometry on the pressure and temperature versus time is theoretically predicted with fairly good agreement with available experimental data. A maximum pressure of 47 MPa and adiabatic temperature of 3500 K are obtained at high compaction, i.e. 50% of the TMD (Theoretical Maximum Density) for stoichiometric mixture, where AIO is shown to be the prevailing gaseous species. At low compaction, we highlight the role of ambient oxygen condition for which the model gives a maximal pressure of 4.2 MPa for Al rich mixtures (stoichiometric ratio of 1.2). (C) 2017 Published by Elsevier Inc. on behalf of The Combustion Institute.