Quantum and neuromorphic computational platforms in integrated photonic circuits require next-generation optical functionalities. Often, increasingly complex on-chip light-routing that allow superpositions not attainable by planar technologies are paramount e.g. for artificial neural networks. Versatile 3D waveguides are achievable via two-photon polymerization (TPP)-based microprinting. Here, a 3D morphology prediction tool which considers experimental TPP parameters, is presented, enabling on-chip 3D waveguide performance simulations. The simulations allow reducing the cost-intensive systematic experimental optimization process. Fabricated 3D waveguides show optical transmission properties in agreement with simulations, demonstrating that the developed morphology prediction methodology is beneficial for the development of versatile on-chip and potentially inter-chip photonic interconnect technology.