Multiscale modeling refers to the combination of multiple models at different scales are used simultaneously to describe a system. These models usually focus on different scales of resolution, atomistic through mesoscopic to continuum (macroscale). Combining the models allows us to understand how microscopic processes governing charge and energy transport influence behaviour, e.g. charge mobilities, at a macroscopic length scale. I will describe our use of random walk Monte Carlo models to address the mesoscopic level and link the length scales. I will show how we can predict the performance of novel optoelectronic devices such as perovskite and organic solar cells and light emitting diodes. In perovskite solar cells, I will demonstrate predictions from a mesoscopic model in which the charge carriers are treated as polarons and that shows the conditions under which hot polarons can beat the Shockley–Queisser limit. Charge carriers in organic semiconductors in organic solar cells and in charge transport layers in perovskite cells are also polarons. Here, polaron transport depends sensitively on molecular packing arrangements. I will describe simulation of charge and energy transport in small molecule and polymer semiconductors with a recently developed fast electrostatics solver and with morphologies obtained with the code Simulation of Atomistic Molecular Structures using an Elastic Network (SAMSEN) developed by my group [1]. Large systems, e.g. 100 polymers each 10 monomers long, can be simulated with SAMSEN in around a day with a desktop computer. [1] A R Smith, I R Thompson, A B Walker J. Chem. Phys. 150, 164115 (2019)