Abstract: Embodiments of the present invention provide apparatuses and methods for a capacity model run-time estimator that determines the transactional capacity of an application. The capacity model run-time estimator is based on a model that includes at least one job with various steps within a run-time window. One or more windows make up shifts and one or more shifts make up a day. Historical data is used to estimate various run-times for each of the steps in a window. If the steps are dependent on transaction characteristics a statistical analysis is used to determine a run-time window estimator model that is dependent on the transaction characteristics. Constraints are placed on the run-time and transaction capacity of the windows, intra-windows, and shifts. The capacity model uses the run-time window estimators and constraints to determine the transaction capacity for an application across various periods and compares the results to forecasted capacity.