Cancer proliferation is driven by a complex interplay of numerous factors, particularly those stemming from mechanical and chemical cell microenvironment play a central role in disease development. Despite their interconnectedness, these factors are often studied in isolation. In-silico data-enhanced approaches constitute a powerful tool for cancer research, as they allow us to integrate different aspects that affect cancer growth within a single model. Here we present a data-enhanced agent-based model integrating tumor metabolism and mechanics to simulate 3D cancer spheroid growth. Our framework simulates metabolic reactions using ordinary differential equations and couples them with a mechanical model, where the interactions between cells and with the extracellular matrix are governed by different forces. The ATP levels, derived from metabolic pathways, regulate cell states (proliferating, quiescent, or apoptotic). These states directly control the mechanical forces exerted on each cell. In parallel, 3D microfluidic experiments were carried out with A549 lung carcinoma cells. The data obtained from these experiments was used for the calibration and validation of our model using Bayesian techniques. Subsequently, the validated model allowed us to simulate tumor spheroid growth under different conditions. By tuning cell motility parameters, the simulations successfully reproduced the motility-driven expansion typical of mesenchymal cells, where high-density extracellular matrices restrict spheroid growth. Conversely, by reducing locomotive forces to mimic epithelial phenotypes, the model captured proliferation-driven growth, where spheroid size becomes less sensitive to matrix density. This study establishes a robust, integrative tool for exploring how metabolic and mechanical cues collectively shape tumor development.