The incorporation of detailed chemistry models in internal combustion engine simulations is becoming mandatory as local, globally lean, low-temperature combustion strategies are setting the path towards a more efficient and environmentally sustainable use of energy resources in transportation. In this paper, we assessed the computational efficiency of a recently developed sparse analytical Jacobian chemistry solver, namely ‘SpeedCHEM’, that features both direct and Krylov-subspace solution methods for maximum efficiency for both small and large mechanism sizes. The code was coupled with a high-dimensional clustering algorithm for grouping homogeneous reactors into clusters with similar states and reactivities, to speed-up the chemical kinetics solution in multi-dimensional combustion simulations. The methodology was validated within the KIVA-ERC code, and the computational efficiency of both methods was evaluated for different, challenging engine combustion modeling cases, including dual fuel, dual direct-injection and low-load, multiple-injection RCCI, direct injection gasoline compression ignition (GDICI), and HCCI engine operation using semi-detailed chemistry representations. Reaction mechanisms of practical applicability in internal combustion engine CFD simulations were used, ranging from about 50 up to about 200 species. Computational performance for both methods was observed to reduce the computational time for the chemistry solution by up to more than one order of magnitude in comparison to a traditional, dense solution approach, even when employing the same high-efficiency internal sparse algebra and analytical formulations. This confirms that consideration of detailed chemistry is not a bottleneck anymore, allowing use of larger and more refined meshes. Further research that focused on algorithms for fast and efficient advection with a large number of species is suggested.