Perovskite/CuIn1 − xGaxSe2 (CIGS) tandem photovoltaics (PV) promises high power conversion efficiencies in combination with the advantages of a light weight and an all thin‐film PV technology. However, the complexity of perovskite/CIGS tandem solar module architectures requires careful optimization of the layer stack under realistic solar irradiation conditions. Here, we provide a systematic numerical study on the energy yield (EY) of perovskite/CIGS tandem solar modules, optimizing the device architecture with regard to irradiance in various climate zones. In particular, variations of the spectral irradiation and the average photon energy are of decisive importance for the location‐specific optimization of the device architecture. Compared with the reference single‐junction CIGS thin‐film PV technology, we demonstrate a strong improvement in EY (>30%) in perovskite/CIGS thin‐film PV for perovskites of a wide range of bandgaps (1.55 ‐ 2.0 eV), reaching up to 52% improvement in EY for the optimal bandgap (around 1.8 eV). Of the two most favored architectures, the two‐terminal and four‐terminal devices, perovskite/CIGS tandem solar modules with low bandgap (∼1.55 eV) perovskite absorbers in the four‐terminal architecture outperform those in the two‐terminal architecture in average by 3.5% relative. However, for wide perovskite bandgaps ranging from 1.75 eV to 1.85 eV, both architectures perform comparably. The improvements in EY for perovskite/CIGS tandem solar modules highlight the potential of this technology but also the vital need for light management in tandem photovoltaics. Energy yield (EY) modelling enables systematic optimizations of the architecture of perovskite/CIGS tandem solar modules with regard to realistic irradiation conditions. Variations of spectral irradiation and the average photon energy are of decisive importance for a location‐specific optimization of the architecture. Compared to a reference single‐junction CIGS solar module, the EY of perovskite/CIGS tandem PV for perovskites of a wide range of bandgaps (1.55 ‐ 2.0 eV) improves significantly by up to 52% for an optimal bandgap of around 1.8eV.