A comprehensive indicator system for digital economy development was constructed, and provincial-level data from 2011 to 2021 were analyzed to explore the linear and nonlinear effects of digital economy on differently-skilled labor demand, with marketization level examined as a mediating variable in the linear pathway analysis and a random forest model was applied to investigate nonlinear relationships. The results show that the digital economy significantly increases high-skilled labor demand while reducing medium-and low-skilled labor requirements, and through marketization mechanisms, it primarily decreases medium-skilled labor demand and elevates high-skilled labor proportion, with nonlinear analysis revealing that high-skilled demand demonstrates initial marginal growth followed by stabilization, while medium-skilled and low-skilled demand undergoes marginal decline with subsequent low-level adjustments. Sectoral variations are also observed, as the industrial sector exhibits greater demand for high-skilled labor, contrasting with the construction sector. While medium-skilled labor shows higher affinity for service and construction industries, and low-skilled labor remains concentrated in agriculture-related sectors, highlighting the technology-driven skill restructuring effect of digital transformation across industries.