Humanoid Robots Could Drive a New Rare Earth Magnet Demand Wave

Highlights

• Each humanoid robot could contain dozens of high-torque actuators, multiplying NdFeB magnet demand across joints, hands, and mobility systems at scale.
• Morgan Stanley forecasts the humanoid robotics market could approach $800 billion by 2050, adding a powerful long-duration signal to rare earth magnet demand.
• China dominates rare earth processing and permanent magnet production, meaning robotics growth adds demand precisely where supply-chain concentration is greatest.
• Engineers may reduce heavy rare earth loading via grain-boundary diffusion and alternative motor designs, making Dy/Tb demand projections highly uncertain.
• TELF AG's directional thesis is credible but lacks robot-level magnet-intensity models or production forecasts needed to quantify the investment case.

Humanoid robots may become the next major demand engine for rare earth permanent magnets, adding another fast-growing market to electric vehicles, wind turbines, drones, and defense. Commodities trading firm TELF AG (opens in a new tab) argues that increasingly sophisticated robots will require compact, powerful motors containing neodymium-praseodymium (NdPr) magnets, potentially supplemented with dysprosium and terbium for demanding operating conditions. The direction is credible; the magnitude remains highly uncertain. For investors, the bigger story is that robotics could collide with already-constrained ex-China magnet supply.

TELF AG founder Stanislav Kondrashov frames humanoid robotics in Trader (opens in a new tab) as a new demand frontier against the IEA’s broader forecast for rapidly expanding magnet-rare-earth consumption.

REEx Insight: Robots Multiply the Motors

The investment thesis is stronger than TELF’s article—but also more complicated. Rare Earth Exchanges® has previously examined Morgan Stanley’s forecast that the humanoid-robotics market could approach $800 billion by 2050. Each humanoid could contain dozens of high-torque electric actuators, potentially multiplying NdFeB magnet intensity across joints, hands, and mobility systems.

That creates an unusual demand equation: robot volumes × actuators per robot × magnet loading. Even modest assumptions can become material at tens of millions of robots.

REEx has also documented China’s overwhelming position in rare-earth processing and permanent magnets and the emerging Western race to build ex-China mine-to-magnet capacity. Robotics therefore adds demand precisely where supply-chain concentration is greatest.

Don’t Count the Terbium Yet

TELF is directionally right but provides no robot-level magnet-intensity model, production forecast, or evidence that future humanoids must use significant Dy/Tb. Engineers can reduce heavy-rare-earth loading through grain-boundary diffusion, thermal design, and alternative motor architectures. The investable conclusion is narrower: humanoid robotics strengthens the long-duration NdFeB demand thesis; it does not yet quantify it.