Assessing ecosystem services (ESs) and exploring the balance between supply and demand is critical for driving urban planning decision-making processes and promoting sustainable land use management, ecological protection, and inhabitants’ well-being. However, integrating and assessing multisource data into spatial planning is challenging due to urban ecosystems’ complexity. This study introduces an innovative methodological approach that connects the power of GeoAI, remote sensing techniques, and Geographic Information Systems to measure and map mismatches between the demand and supply of ESs, providing a new perspective on urban planning. In developing a Google Earth Engine cloud computing platform, a Random Forrest (RF) classifier is employed to process and automatically classify multispectral satellite images and interpret complex Land Use Land Cover (LULC) patterns. This enables an accurate mapping of ESs supply-demand mismatch. The RF classifier accuracy was rigorously measured by comparing the model predictions with those derived from the visual interpretation of high-resolution Google Earth images. Overall Accuracy (OA) and the Kappa coefficient were the accuracy metrics. The results of our study reveal a significant spatial mismatch between ES demand and supply across different land use types, shedding light on urban areas where ESs are either over-utilized or under-provided. This enlightening information can guide future urban planning decisions. Following a performance-based approach, this study demonstrates significant utility in aiding policymakers and urban planners in making informed decisions for enhancing urban sustainability.