Actin polymerization drives cell membrane protrusions and the propulsion of intracellular pathogens. The molecular mechanisms driving actin polymerization are not yet fully understood. Various mathematical models have been proposed to explain how cells convert chemical energy released upon actin polymerization into a pushing force on a surface. These models have attempted to explain puzzling properties of actin-based motility, including persistent attachment of the network to the membrane during propulsion and the interesting trajectories of propelled particles. These models fall generally into two classes: those requiring filament (+)-ends to fluctuate freely from the membrane to add subunits, and those where filaments elongate with their (+)-ends persistently associated with surface through filament end-tracking proteins ("actoclampin" models). This review compares and contrasts the key predictions of these two classes of models with regard to force-velocity profiles, and evaluates them with respect to experiments with biomimetic particles, and the experimental evidence on the role of end-tracking proteins such as formins and nucleation-promoting factors in actin-based motility.