We introduce a new prediction-oriented method for hazard assessment of volcaniclastic debris flows, based on multiple models. We apply our procedure to a case study of the 1955 Atenquique flow, using three widely used depth-averaged models. Depending on how it is looked at, the exercise provides useful information in either model selection or data inversion. Connecting inverse problems and model uncertainty represents a fundamental challenge in the future development of multi-model solvers.