Architectures with various degrees of integration are investigated for water splitting devices using the energy of light for fuel production. The many approaches presented in literature for such ‘photo driven catalytic (PDC) devices’ are reviewed and discussed in perspective of their scalability to large area. Then, back-of-the-envelope type techno-economic considerations for such systems are presented. Compared to the benchmark, consisting of large electrolyzers coupled to the grid, it was found that PDC devices could be competetive in places with high irradiation, given the assumption that no compromises on system stability have to be made compared to stand-alone PV-systems for electricity generation. In agreement with literature, it was found that the cost of the PV part dominate the hydrogen generation costs, based on today's technology. Thus, device architectures that allow low cost PV (by e.g. avoiding use of costly materials or introducing further inherent loss mechanisms) are considered the most promising ones.