<p>Implantable biohybrid systems with computer-controlled actuation offer the capacity to modulate biological forces, but require biocompatible, self-sustaining, and scalable actuators. Engineering biological muscles can fulfill this need. However, muscle fatigue limits the fundamental capabilities of muscle-actuated systems. Here we present a fatigue-resistant myoneural actuator (MNA) with engineered recruitment biophysics in a rodent model. The MNA is based on manipulating native axonal composition through sensory reinnervation. This regenerative approach establishes functional neuromuscular junctions and redirects volitional control to computer control via nerve stimulation while maintaining self-sustainability. Compared to native muscles without the myoneural manipulation, fatigue resistance is augmented by 260%. Furthermore, we demonstrate closed-loop control with reversible neural isolation of the actuator, preventing unintended neural signaling to the central nervous system during operation. To illustrate the potential of the MNA technology, we present a biohybrid neuroprosthetic interface and a biohybrid organ system capable of modulating neural afferents and organ mechanics, respectively. Our framework demonstrates augmented biological muscle actuation while maintaining inherent tissue properties, bridging the technological gap for implantable biohybrid systems.</p>