This paper constructs an evaluation model of China’s high-quality economic development based on deep learning methods, and applies kernel density estimation and Markov chains to clarify the spatiotemporal dynamic evolution and spatial development trends of high-quality economic growth. The study reveals the following findings: First, it can be observed from the changes in the color of the heat map that, over time, the high-quality economic development index of most cities shows an upward trend; however, differences persist in the growth rates and volatility among cities. Second, a comparative analysis between modern and traditional methods (entropy weight method and equal weighting method) indicates that the high-quality economic development index calculated using deep learning outperforms traditional linear evaluation approaches. Third, based on the kernel density estimation graphs (which can reveal the trends of the dynamic evolution of high-quality economic development in various regions), the research results show that the levels of high-quality economic development in China’s four major regions (eastern, central, western and northeastern) have improved to varying degrees over time. Fourth, the spatial evolution trend of China’s high-quality economic development, revealed through Markov chain analysis, shows that China’s high-quality economic development is significantly influenced by spatial spillover effects from neighboring areas. However, the direction and intensity of this influence depend on the economic development gaps between regions. When these gaps are small, there is a significant mutual promotion effect; but when the gaps become too large, the driving effect of high-level regions on low-level regions is constrained.