Excitons, or coupled electron-hole excitations, are important both for fundamental optical properties of materials as well as and for the functionality of materials in opto-electronic devices. Depending on the material they are created in, excitons can come in many forms, from Wannier–Mott excitons in inorganic semiconductors, to molecular Frenkel or bi-molecular charge-transfer excitons in disordered organic or biological heterostructures. This multitude of materials and exciton types poses tremendous challenges for ab initio modeling. Following a brief overview of typical ab initio techniques, we summarize our recent work based on Many-Body Green’s Functions Theory in the GW approximation and Bethe–Salpeter Equation (BSE) as a method applicable to a wide range of material classes from perfect crystals to disordered materials. In particular, we emphasize the current challenges of embedding this GW-BSE method into multi-method, mixed quantum-classical (QM/MM) models for organic materials and illustrate them with examples from organic photovoltaics and fluorescence spectroscopy. Our perspectives on future studies are also presented.