The functionality of photo-driven electrochemical devices relies on complicated and coupled multi-physics processes, taking place on multiple temporal and spatial scales. Device modelling actively and efficiently supports the choice of the most interesting conceptual design pathways, material choices, and operating approaches—in terms of efficiency, cost, robustness, scalability, practicability, and sustainability. Models of photo-driven electrochemical devices and components incorporate a wide range of physical phenomena and complexity. In this chapter, we discuss the merit of and lessons learned from device models incorporating different degrees of complexity and dimensionality. This includes (i) steady state and transient zero-dimensional models for the discussion of material choices, degradation, and device integration, (ii) steady state one-dimensional models in order to understand the role of the material choice and semiconductorelectrolyte interface on the performance of photocatalytic and photoelectrochemical device designs, and (iii) steady state two-dimensional models for a discussion of the effect of temperature and heat management on performance. We show that models with different complexities, in terms of dimensionality and inclusion of physical phenomena, can provide answers to different research questions, and each have merit when it is ensured that the underlying assumptions are well known.