A multiscale, hierarchical computational framework is presented for modeling homogeneous-heterogeneous reactors, which exhibit a large disparity in length and time scales. Scales range from quantum, to atomistic, to mesoscopic, to macroscopic. The coupling mechanisms between scales are discussed and illustrated with examples from CO and CH4 oxidation on platinum. Estimation of reaction mechanism parameters, based on first principle quantum calculations and semi-empirical techniques, is briefly reviewed. These kinetic mechanisms are key input into molecular, continuum, or mesoscopic models. Some emphasis is placed on surface diffusion, which typically falls outside the realm of atomistic models, but it can affect reaction rates and pattern formation on catalytic surfaces. An efficient methodology for parameter optimization of multiscale models is also presented. Finally, we show how mesoscopic models constitute a promising alternative to atomistic Monte Carlo (MC) simulations to account for intermolecular forces, which cannot be properly captured through continuum, mean field (MF) models. Application of these mesoscopic theories to microporous catalysts, such as zeolites, is also discussed.