Abstract: Analyzing the carbon emission patterns of agriculture in the interlaced areas of agriculture and animal husbandry is of great value for promoting the low-carbon transformation and sustainable development of regional agriculture. Taking Jalaid Banner, Inner Mongolia as the case study area, the study employed an agricultural carbon emission measurement model and, based on factor drivers, constructed a suite of models (including STIRPAT-LMDI driver decomposition model, Tapio-LMDI economic decoupling model, and STIRPAT-SAM development prediction model) to conduct a comprehensive analysis of the evolution characteristics, driving factors, economic decoupling status, and development trends of agricultural carbon emissions. The results showed that during 2010-2023, the temporal variation of agricultural carbon emissions was divided into three stages: the slow growth stage, the high-speed growth stage, and the "J-curve" decline stage. Carbon emissions mainly originated from chemical fertilizer application, agricultural production electricity consumption, and agricultural diesel input, combined with rice cultivation and livestock farming (cattle, sheep, and swine), collectively forming a dual-drive mechanism of "arable land utilization-livestock breeding". The carbon emission effects of the driving factors, including population size, economic development, technological progress, population structure, economic structure, technological investment, capital investment, and consumption level, were 211 593, 109 846, −188 810, 212 351, 127 249, 163 060, 135 846, and 182 823 tons, respectively. Meanwhile, the synergistic effect of various driving factors dynamically determined the evolution direction of the decoupling index, leading to two categories (decoupling and negative decoupling) and four states (strong decoupling, weak decoupling, expansive negative decoupling, and strong negative decoupling) in the relationship between agricultural carbon emissions and economic growth. Projections for 2024-2030 under different development scenarios, considering changes in driving factors, revealed that carbon reduction effects ranked as follows: coordinated scenario > low-carbon scenario > planned scenario > baseline scenario. This study elucidates the fundamental patterns of agricultural carbon emissions in Jalaid Banner while validating the innovative utility of synergistic modeling. The agricultural carbon emission patterns in Jalaid Banner demonstrate significant correlations with regional resource endowment and agricultural development models, suggesting that emission reduction strategies should integrate regional characteristics with the driving factors to promote coordinated development of agricultural carbon emission reduction and economic growth.