What if a robot could rethink its own morphological representation to better meet the demands of diverse tasks? Most robotic systems today treat their physical form as a fixed constraint rather than an adaptive resource, forcing the same rigid geometric representation to serve applications with vastly different computational and precision requirements. We introduce MorphIt, a novel algorithm for approximating robot morphology using spherical primitives that balances geometric accuracy with computational efficiency. Unlike existing approaches that rely on either labor-intensive manual specification or inflexible computational methods, MorphIt implements an automatic gradient-based optimization framework with tunable parameters that provides explicit control over the physical fidelity versus computational cost tradeoff. Quantitative evaluations demonstrate that MorphIt outperforms baseline approaches (Variational Sphere Set Approximation and Adaptive Medial-Axis Approximation) across multiple metrics, achieving better mesh approximation with fewer spheres and reduced computational overhead. Our experiments show enhanced robot capabilities in collision detection accuracy, contact-rich interaction simulation, and navigation through confined spaces. By dynamically adapting geometric representations to task requirements, robots can now exploit their physical embodiment as an active resource rather than an inflexible parameter, opening new frontiers for manipulation in environments where physical form must continuously balance precision with computational tractability.