Planting oat forage in fallow fields during winter and producing total mixed ration (TMR) silage can effectively address issues of land wastage and forage shortages while maintaining forage quality. This study used oats and common vetch grown in winter fields in southern China as base materials, with additives including corn flour, soybean meal, corn lees, cottonseed meal, and premixes to formulate mixed feeds with roughage-to-concentrate ratios of 75:25, 70:30, and 65:35 on a dry matter basis. TMR silage was inoculated with a customized mixed lactic acid bacteria (LAB) additive composed of Lactobacillus plantarum 160 (patent number ZL202210218695.5), Lactobacillus pentosus 260 (patent number ZL202210204293), and Lactobacillus buchneri 225 (patent number ZL202210204293), at a ratio of 2:1:1, with addition rates of 4 × 106, 2 × 106, and 2 × 106 cfu/g, respectively (IN), while sterile distilled water served as the control (CK). After a 60-day fermentation, the cornell net carbohydrate protein system (CNCPS) and in vitro digestion analysis were used to assess the effects of different roughage-to-concentrate ratios on the carbohydrate and protein components and ruminal degradation rate of fermented TMR (FTMR) silage, as well as to evaluate the impact of mixed LAB inoculation on FTMR nutritional quality, fermentation quality, and aerobic stability. The results indicated the following: (1) Regardless of the LAB addition, dry matter (DM), ether extract (EE), crude protein (CP), and Ash contents significantly decreased (p < 0.05) as the concentrate level decreased. In the IN group, as the concentrate level decreased, the water-soluble carbohydrate (WSC) content significantly increased (p < 0.05), the pH significantly decreased (p < 0.05), and the NH3-N/TN significantly decreased (p < 0.05), with LAB counts significantly higher at a 65:35 roughage-to-concentrate ratio than in the other two groups. In the CK group, no significant changes (p > 0.05) were observed in the WSC content, pH, or LAB counts. (2) CNCPS analysis showed that in the IN group, the carbohydrate (CHO) content at a 75:25 roughage-to-concentrate ratio was significantly higher than in the other two groups (p < 0.05), while the non-utilizable carbohydrate (CC) content was significantly lower (p < 0.05). As the concentrate levels decreased, the non-protein nitrogen (PA) and moderately degradable true protein (PB2) content significantly increased (p < 0.05), whereas the rapidly degradable true protein (PB1) and slowly degradable true protein (PB3) content significantly decreased (p < 0.05). In the CK group, the CHO, PA, PB2, and PC content significantly increased (p < 0.05) as concentrate levels decreased, while the PB1 and PB3 content significantly decreased (p < 0.05). (3) In vitro digestibility characteristics indicated that gas production (GP) in the IN group was significantly lower than in the CK group (p < 0.05), with crude protein degradability increasing as concentrate levels decreased, regardless of the LAB addition. (4) At a 65:35 roughage-to-concentrate ratio, aerobic stability in the IN group was significantly higher than in the CK group (p < 0.05). In conclusion, higher concentrate ratios in total mixed rations (TMRs) with varying roughage-to-concentrate proportions improve the nutritional quality and promote the ruminal degradation of the FTMR. LAB inoculant addition could be an effective approach for addressing FTMR feed challenges.