Multiple myeloma (MM) remains incurable despite the abundance of novel drugs. As it has been previously shown, preclinical 2D models fail to predict disease progression due to their inability to simulate the microenvironment of the bone marrow. In this review, we focus on 3D models and present all currently available ex vivo MM models that fulfil certain criteria, such as development of complex 3D environments using patients’ cells and ability to test different drugs in order to assess personalized MM treatment efficacy of various regimens and combinations. We selected models representing the top-notch ex vivo platforms and evaluated them in terms of cost, time-span, and feasibility of the method. Finally, we propose where such a model can be more informative in a patient’s treatment timeline. Overall, advanced 3D preclinical models are very promising as they may eventually offer the opportunity to precisely select the optimal personalized treatment for each MM patient.