Abstract: The exquisite astrometric dataset that Gaia DR3 offers to the scientific community makes the XMC a unique test laboratory to investigate and refine methodologies and models. Classification strategies based on Neural Networks allow to select LMC and SMC samples with less than 10% contamination from the Milky Way (Jiménez-Arranz+23a,b). With this dataset, the LMC is the first galaxy for which we can map the 3D velocity field across the whole disc (Jiménez-Arranz+23a). However, line-of-sight velocities, the third component of the stellar motion, are only available for a small subset of the current Gaia data, limiting studies of the kinematics perpendicular to the LMC disc plane. We synergise new SDSS-IV/V line-of-sight velocity measurements with existing Gaia DR3 data, increasing the 5D phase-space sample by almost a factor of three. Using this unprecedented dataset, in this talk we interpret and model the vertical structure and kinematics across the LMC disc (Jiménez-Arranz+25).