We present a systematic approach to commissioning of the compensator-based IMRT in Pinnacle treatment planning system for commercially manufactured brass compensators. Some model parameters for the beams modulated by the variable-thickness compensators can only be associated with a single compensator thickness. To intelligently choose that thickness for beam modeling, we empirically determined the most probable filter thickness occurring within the modulated portion of the compensators typically used in clinics. We demonstrated that a set of relative output factors measured with the brass slab of most probable thickness (2 cm) differs from the traditionally used open field set, and leads to improved agreement between measurements and calculations, particularly for the larger field sizes. By iteratively adjusting the modifier scatter factor and filter density, the calculated effective attenuation of the flat filters was brought to within 2% of the ion chamber measurement for the clinically-relevant range of filter thicknesses, depths and filed sizes. Beam hardening representation in Pinnacle provides for adequate depth dose modeling beyond the depth of about 5 cm. Disagreement at shallower depth for the large field sizes is likely due to the algorithm's inability to account for the low-energy scattered photons generated in the filter. The average ion chamber point dose error at isocenter for ten clinical compensator-based IMRT plans was under 1%. A biplanar 3D diode dosimeter was calibrated and validated for use with the compensators. The average gamma analysis (3%/3 mm) passing rate for ten IMRT plans was 98.9% ± 1.0%. The device is particularly attractive because it easily generates dose comparisons at both the fraction and beam levels. Overlaying dose profiles for individual beams would easily uncover any errors in compensator orientation.