First-principles modeling, based on the direct simulation Monte Carlo (DSMC) method, is applied to quantitatively explain the physics of stainless-steel (SS316L) vapor plumes and jets and powder particle entrainment under conditions relevant to laser welding and laser powder bed fusion additive manufacturing. The DSMC is a particle-based gas kinetic simulation approach, with few built-in assumptions, that resolves the Knudsen and diffusion layers in vapor microjets. The simulations are performed for different background gas species (argon, helium, and neon) at background gas pressures from 0.1 to 5 bar and molten pool temperatures from 3250 K to 4000 K. The simulations capture the evolution of the vapor jet structure from unsteady to steady-state vapor flow, the transition from subsonic flow at low surface temperatures to supersonic flow at surface temperatures close to boiling, as well as the effects of the background gas species and pressure. The simulations show that the vapor jet structure, as well as the magnitude and direction of the entrained gas close to the laser spot, greatly affects the process environment, creating particle spatter and denudation around the laser track. High pressure and lighter background gas can decrease the amount of spatter. At high pressure, the evaporation is reduced, which forms a low-velocity subsonic plume. For lighter gas species, the gas-entrainment velocity increases while the jet radius decreases which forces entrained particles to merge with the melt pool. The powder particle simulations reveal that the width of the denudation zone strongly depends on particle diameter and increases at low pressure and lighter background gas. This study emphasizes that the ambient gas species and the pressure in the build chamber should be treated as process parameters that are as important as controlling laser power and scan speed when trying to prevent defects. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract DE-AC52-07NA27344. Lawrence Livermore National Security, LLC. LLNL-JRNL-829362. A.N.V also acknowledges supported by the National Science Foundation, USA through RII-Track-1 Future Technologies and Enabling Plasma Processes project (award OIA-2148653).