Abstract Facing the conflicts of climate change, energy consumption, carbon emission, and economic development, it is essential to investigate the impacts of the carbon tax policy implemented in specific regions. A CGE-based multi-dimensional carbon policy (CMDCP) model is developed to i) explore the inter-provincial interdependences by interfering with the economic policies of a single province, and ii) quantify interactive relationships among various components including climate, energy, carbon economy and tax. Integrated approach of computable general equilibrium model and input-output analysis is applied to a series of segmented carbon tax schemes for Guangdong IC modeling and China IE modeling. It is found that when the carbon tax rate is 100 yuan/ton, the GDP of Guangdong will fall by less than 0.5% under three scenario types. At the same time, they could bring 1.3, 1.2 and 1.6 million tons of emission reductions. Levying the carbon tax based on the difference in carbon emission volume is most beneficial for emission intensity reduction. For China, the impact of the segmented carbon tax in specific province has a slight impact on the entire supply chain emissions. It is suggested that a carbon tax of 10–40 yuan/ton could be adopted by Guangdong. Moreover, Guangdong could consider implementing the stepped carbon tax for it can effectively avoid the lack of flexibility of traditional carbon tax policy.