The demand for greater computing power to solve challenging problems in science, medicine, and economics continues to rise at a rapid pace. Future computers based on quantum bits (qubits) will exponentially speed up certain types of problems related to information security, data search, drug design, and quantum chemistry. Although this speedup increases exponentially with the number of qubits, the energy consumption is expected to only increase at most linearly. Recently, a qubit candidate with remarkable properties was experimentally demonstrated: the TbPc2 single-molecule magnet, which features a well isolated nuclear spin as the qubit combined with the attractive and unusual property of electrical controllability. The TbPc2 molecule stands out among other qubit candidates because it brings together the best of all worlds: long-lived qubit coherence, strong coupling, and fast controllability. We will develop a multiscale modeling approach that ranges from different levels of ab-initio quantum chemistry simulations to effective models and time-dependent external control, and will use this approach for the systematic design and control of quantum information processors with TbPc2 single-molecule magnets. The impact of this work will be two-fold: (i) We will develop new quantum chemistry simulation techniques capable of treating multiscale problems with significant multi-reference character, such as the TbPc2 molecule on a substrate, which are not currently amenable to existing methods, and (ii) we will design quantum information processors based on these molecules and our simulation results, and we will develop robust control methods to operate them. The developed techniques will be general and transferable to other molecules and will thus more broadly impact the field of quantum chemistry with their integration into open-source codes.