Sub-GeV dark matter (DM) particles produced via thermal freeze-out evade many constraints on heavier DM candidates but face new constraints from experiments, astrophysics, and cosmology. This work combines these constraints to perform frequentist and Bayesian global analyses of fermionic and scalar sub-GeV DM coupled to a dark photon with kinetic mixing. For fermionic DM, viable parameter regions near the dark photon resonance are found, expanding with particle-antiparticle asymmetry. Scalar DM has a velocity-dependent annihilation cross section that evades strong constraints. Bayesian model comparison shows asymmetric fermionic DM and symmetric scalar DM are preferred over symmetric fermionic DM due to reduced fine-tuning. Future experiments may find benchmark scenarios already in tension with current constraints, suggesting a new benchmark point. Key findings include the importance of the resonance parameter ε_R and the role of DM self-interactions in astrophysical systems. The analysis highlights the need for a comprehensive parameter space exploration and the use of global fitting frameworks like GAMBIT. The results show that sub-GeV DM models must satisfy stringent constraints from CMB and X-ray observations, with the resonance parameter playing a central role in determining the viability of different models. The study also emphasizes the importance of considering DM asymmetry and the impact of DM self-interactions on astrophysical observations. The results suggest that future experiments should target new benchmark points consistent with current constraints.Sub-GeV dark matter (DM) particles produced via thermal freeze-out evade many constraints on heavier DM candidates but face new constraints from experiments, astrophysics, and cosmology. This work combines these constraints to perform frequentist and Bayesian global analyses of fermionic and scalar sub-GeV DM coupled to a dark photon with kinetic mixing. For fermionic DM, viable parameter regions near the dark photon resonance are found, expanding with particle-antiparticle asymmetry. Scalar DM has a velocity-dependent annihilation cross section that evades strong constraints. Bayesian model comparison shows asymmetric fermionic DM and symmetric scalar DM are preferred over symmetric fermionic DM due to reduced fine-tuning. Future experiments may find benchmark scenarios already in tension with current constraints, suggesting a new benchmark point. Key findings include the importance of the resonance parameter ε_R and the role of DM self-interactions in astrophysical systems. The analysis highlights the need for a comprehensive parameter space exploration and the use of global fitting frameworks like GAMBIT. The results show that sub-GeV DM models must satisfy stringent constraints from CMB and X-ray observations, with the resonance parameter playing a central role in determining the viability of different models. The study also emphasizes the importance of considering DM asymmetry and the impact of DM self-interactions on astrophysical observations. The results suggest that future experiments should target new benchmark points consistent with current constraints.