The force produced by activated skeletal muscle increases considerably (i.e., about 1.5 – 2 times) upon stretching its half sarcomeres a small amount (i.e., about 10 nanometers), and plateaus upon further stretching. Similar force transients can be observed across a broad range of stretch velocities, and across a broad range of species. The initial rise in force is known as short-range-stiffness and is believed to help stabilize animal movements by reducing the need for neural control. Short-range-stiffness is difficult to assess during movements but could be predicted using computer simulations with musculoskeletal models. However, current simulations rely on phenomenological Hill-type muscle models, which cannot explain history-dependent changes in short-range-stiffness with prior shortening, that may occur during movements. In contrast, simple biophysical muscle models have history dependence, but cannot reproduce force transients during stretch. This may be because such simple models typically only have one bound and one unbound crossbridge state, while experimental data suggest that crossbridges have at least two bound states (e.g., weakly or strongly) and two unbound states (e.g., detached or forcibly detached). Complex biophysical models including such states have previously been shown to fit force transients better, but at a high computational cost. To facilitate implementing such models in musculoskeletal simulations, we therefore applied a moment approximation to their crossbridge distributions and greatly reduced their computational cost. These models reproduced both force transients and typical steady-state force-velocity relations for various muscles. Ultimately, we will implement biophysical muscle models in musculoskeletal simulations to elucidate how muscle properties shape movement biomechanics.