CO2 foam fracturing can reduce reservoir damage and contribute to the stimulation of unconventional oil and gas reservoirs. However, there needs to be more quantitative methods to evaluate the influence of the CO2-H2O ratio of foam fracturing fluid on the conductivity, and the optimization of foam fracturing fluid system is insufficient. To select the CO2 foam fracturing fluid system suitable for the conglomerate reservoir in the Mahu basin, proppant embedment experiments are carried out with the treatment of foam fluid with different CO2-H2O ratios at simulated reservoir conditions, and each interval of embedment depth is obtained. A conductivity model considering the heterogeneous proppant embedment is established to calculate the effects of different foam fluids on improving the conductivity and analyze its mechanism. The results show that using CO2 to replace part of the water-based fracturing fluid can reduce the degree of proppant embedment. With the increase of the CO2-H2O ratio, the effect of improving the conductivity weakens and reaches the upper limit gradually, increasing by about 12% compared with that of water-based fracturing fluid. When the ratio is 7:3, the conductivity of samples with low clay content (< 20%) increases to the upper limit. However, the conductivity of samples with high clay mineral content (≥20%) is more sensitive to the influence of the CO2-H2O ratio, and the c upper limit is lower. CO2 foam fracturing fluid can improve the hydrological capacity of the propped fracture-formation system, but its effect has been underestimated in this experiment. This study can reference the CO2-H2O ratio optimization of the CO2 foam fracturing fluid system.