Acoustic graphene plasmons (GPs) exhibit an exceptional density of electromagnetic states within the mid-infrared (MIR) and terahertz frequency ranges, leading to a pronounced near-field amplification and localization at the nanometer scale. This characteristic renders them highly promising for the development of ultra-sensitive plasmon-enhanced MIR sensing and spectroscopic applications. However, the tight spatial confinement inherent to acoustic GPs results in a significant momentum mismatch, which in turn leads to poor coupling efficiency with light in free space. To overcome this limitation, we leverage numerical simulations to show that GPs can act as an intermediary to facilitate efficient excitation of acoustic GPs by incident MIR radiation, achieving an extinction spectrum peak value of approximately 0.9. The proposed gas sensor based on the acoustic GP resonator is composed of pristine, large-area graphene, an array of periodic metal nanocubes, and a slender polyethylenimine (PEI) layer that adsorbs CO2, situated between the nanocubes and the graphene sheet. The sensing performance of the proposed sensor is numerically investigated. It is demonstrated that by incorporating a CO2-absorbing material into the acoustic GPs, we can perform highly sensitive assessments of the absorption bands within the PEI layer upon CO2 adsorption. The observed behavior of the acoustic GPs indicates a weakening and broadening with increasing CO2 concentrations, a phenomenon attributed to the alteration in the permittivity of the PEI in the interstitial layer due to CO2 adsorption. Numerical simulation results indicate that the sensitivity of the proposed hybrid gas sensor can reach up to 0.0183% ppm−1, which offers a remarkable 95-fold enhancement over the gas sensor based on graphene nanoribbons. Our findings underscore the potential of acoustic GP nanoresonators as a transformative platform for ultrasensitive plasmon-enhanced sensing applications, particularly when integrated with various gas adsorption layers or molecular agents.