A better understanding of the greenhouse gas (GHG) emissions intensity of upgrading oil sands bitumen is needed to inform industry and government decisions related to climate change mitigation. We develop an enhanced version of the Oil Sands Technologies for Upgrading Model (OSTUM 2.0), a life cycle-based model that estimates energy use and GHG intensities of upgrading technologies at the process unit level. We apply OSTUM 2.0 to commercial upgrading technologies operating in Alberta, Canada, and propose baseline estimates and ranges of direct and indirect GHG intensities for delayed coking based- (DC), hydroconversion based- (HC), and combined hydroconversion and fluid coking based upgrading (HC/FC). We identify potential drivers of variability in upgrading GHG intensity. These include the application of different upgrading technologies (up to 36% variation in baseline GHG intensities), variation in the properties of upgrading products (up to 91% variation in baseline GHG intensities, a factor not previously modeled in the literature), and changes in the energy efficiency of operations and use of byproducts as fuels (up to 45% variation in baseline GHG intensities). OSTUM 2.0 improves upon life cycle-based literature by facilitating more detailed modeling, assessment, and comparison of the GHG intensity of commercial upgrading schemes, using consistent boundaries, assumptions, public data sources, and calculation methods. The systems-level modeling approach allows comprehensive characterization of upgrading GHG emissions and drivers of variability. Our findings suggest that modeling different technology configurations, product properties, and fuel mixes can increase the representativeness of upgrading GHG intensity estimates.