Tesla CEO Elon Musk said in September 2026 that the motor in its Cybercab robotaxi will use no rare earths while maintaining driving range, according to media reports cited by Chinese outlet Huxiu. The announcement also mentioned that the electric drive unit is 18% smaller and 25% lighter. But for motor engineers, the claim raises more questions than it answers, the report notes.
The comparison baseline, output power at specific speeds, sustained torque, temperature rise, efficiency and noise all need scrutiny before concluding that dropping rare earths makes motors better, according to the report. Auto Business Review, cited by Huxiu, said Musk's statement is a corporate performance claim, not a conclusion from independent comparative testing. The real question is how much room exists to replace rare-earth dependence in EV traction motors.
Rare-earth-free motors are already in mass-produced vehicles. The debate has moved beyond whether it is possible. The key issues are whether savings on magnet procurement translate into added costs for bearings, gearboxes and cooling systems, and whether similar peak power can deliver similar sustained output and driving experience.
Multiple technical paths
Motors convert electrical energy into mechanical energy through magnetic fields. Permanent magnets provide fields, but so do energized coils, and reluctance torque from rotor geometry differences can also produce torque without permanent magnets or excitation windings. Thus, rare-earth-free motors follow several routes: electrically excited synchronous motors use rotor coils to build fields; induction motors use electromagnetic induction; synchronous reluctance motors exploit rotor reluctance differences; and permanent magnet motors can use rare-earth-free magnets.
Rare-earth permanent magnets offer strong fields in small volumes, enabling compact, efficient motors. But magnet performance does not directly equal motor performance. Magnetic circuit design, current loading, reluctance torque and cooling all affect the final result. Understanding alternatives requires distinguishing three levels: heavy-rare-earth-free, rare-earth-free, and permanent-magnet-free. Heavy-rare-earth-free means removing elements like dysprosium and terbium while possibly retaining light rare earths such as neodymium and praseodymium. Rare-earth-free does not necessarily mean no permanent magnets—ferrite magnets without rare earths are an option, though some high-performance ferrites use lanthanum modification. Permanent-magnet-free designs change the rotor entirely.
BMW's fifth-generation electric drive uses electrically excited synchronous motors, avoiding rare-earth permanent magnets, according to a 2021 BMW announcement cited by Huxiu. The iX3 has used this motor since 2020, and it also powers the iX and i4. Such designs must solve power supply to rotating rotors and rotor heat dissipation. ZF's inductive power supply through the rotor shaft is one approach.
Another direction combines permanent magnet torque with reluctance torque. Ferrite-assisted synchronous reluctance motors are not simple magnet swaps; they require rotor structure and magnetic circuit design so both torque types share output. These routes are all discussable, but none can be confirmed for the Cybercab based on public information. Tesla's 2023 plan for a next-generation rare-earth-free permanent magnet motor provides context, but not confirmation of the Cybercab's actual configuration and materials.
Numbers need context
The reported 18% size reduction and 25% weight reduction refer to the electric drive unit, which may include components beyond the motor itself. The comparison scope and output conditions are unspecified. Housing, cooling, transmission and system integration all affect results. Similarly, "maintaining range" is a vehicle-level claim involving battery capacity, weight, aerodynamics, tires and drive efficiency—not a direct measure of motor efficiency change.
Motor power, torque and speed follow the formula: power (kW) = torque (Nm) × speed (rpm) ÷ 9549.3. For typical radial-flux motors, torque roughly scales with rotor effective volume under similar electromagnetic loading, but across different materials, cooling and structures, volume alone cannot predict torque. Increasing speed improves power density but not necessarily torque density. A hypothetical example: if a design delivers 70% of baseline torque at the same volume but maintains that torque at double the speed, output power reaches 140% of baseline. Real motors face voltage, current, flux-weakening, mechanical strength and thermal constraints at high speeds. Maximum torque and maximum speed rarely occur at the same operating point.
Proterial (formerly Hitachi Metals) published technical data on July 24, 2023, comparing a ferrite prototype motor against a benchmark automotive neodymium-iron-boron permanent magnet motor, according to Huxiu. The design fixed rotor and stator diameters and considered operating temperature and high-speed strength, but did not disclose the benchmark motor's brand, model or full test conditions. The prototype had about 7.3% lower power, 37% lower maximum torque, and 50% higher maximum speed rating. These are design-specific results, not universal gaps for ferrite motors.
A 37% lower maximum torque does not mean 37% weaker launch performance in the vehicle. Low-speed torque curves, gear ratios, power supply capability and tire traction all matter. Increasing the gear ratio can boost wheel torque but requires matching motor speed range, top speed, gear loads and transmission losses. Higher speeds also impose new demands on rotor strength, balance, bearings and lubrication, with potential costs in NVH (noise, vibration and harshness) and thermal management. Speed changes shift electromagnetic and mechanical excitation frequencies; if they approach structural resonances, vibration or whine can amplify. The Cybercab's reported weight, size and range data are insufficient to assess these acoustic issues or any added isolation and noise-reduction investments.
Thermal management is similarly complex. Higher speeds may increase iron losses, AC winding losses and mechanical losses, but final temperature rise depends on total losses and cooling. One cannot simply assert high-speed motors run hotter, nor assume rotors without excitation windings generate no losses. The Astemo case cited in the report illustrates the trade-offs: according to an October 2025 Astemo announcement, its ferrite-magnet-assisted synchronous reluctance main drive motor achieves 180 kW output but is about 30% larger than its conventional rare-earth permanent magnet motor. To control heat, the design uses oil-immersed cooling for the winding slots and ends.
