Oppenheimer analyst Colin Rusch said Tesla is unlikely to generate meaningful sales revenue from its Optimus humanoid robot before 2029, pushing his Optimus ramp assumptions out by two quarters. Following a visit to Tesla's Austin operations, Rusch said the company pointed to finalizing the hardware design for Optimus's initial ramp and is targeting first commercial revenue in late 2027, but he wrote that Street expectations for humanoid revenue scale-up are overly optimistic, citing hardware and operating software complexity and supply chain preparedness. Rusch, who rates Tesla at Perform, was more upbeat on other parts of the business, saying the Cybercab is ramping and driving manufacturing costs materially lower, with a modular design enabling parallel production and up to 40% mass reduction, and describing Tesla's 50% to 60% robotaxi cost advantage as becoming a reality. He also flagged Tesla's push into semiconductor manufacturing as bold but critically important, with a pilot facility tracking toward a $25 billion-plus capital spending target. Oppenheimer kept its 2026 revenue estimate at $105 billion but trimmed 2027 to $108.8 billion, reflecting the slower Optimus ramp offset by faster Cybercab production.
Morgan Stanley Flags Deepening SpaceX-Tesla AI Ties Ahead of SpaceX IPO
Morgan Stanley has identified expanding technological and operational links between SpaceX and Tesla, the two Elon Musk-led companies pursuing artificial intelligence applications in robotics, energy and manufacturing. In a note led by analyst Adam Jonas, the bank said the companies share tech, talent and infrastructure in a shared mission of converting energy to intelligence, and it cited SpaceX President Gwynne Shotwell saying there are synergies between Tesla and SpaceX and a convergence in what both are trying to accomplish. The bank counted 71 references to Tesla in SpaceX's S-1 registration filing, of which 34% concerned direct collaboration between the companies and 41% related to governance matters and transactions between related parties. Areas of cooperation include Terafab, a semiconductor manufacturing project being developed in Texas with Intel whose total construction cost has reportedly been estimated at more than $119 billion, and joint work on an AI agent platform called Macrohard. SpaceX also buys batteries and vehicles from Tesla, including $506 million in Megapack energy storage systems and $131 million in Cybertrucks during 2025 to support the Colossus computing cluster, while both companies target annual vertically integrated solar manufacturing capacity of 100 gigawatts in the United States. Morgan Stanley described a reciprocal arrangement in which SpaceX supplies Tesla with computing resources and satellite connectivity while Tesla contributes robotics, energy storage and manufacturing expertise, with SpaceX computing capacity projected to reach 4.9 gigawatts by the end of 2027 and terrestrial capacity expected to rise to 15.3 gigawatts by 2031. Tesla is integrating Starlink satellite connectivity into its vehicle range beginning with Cybercab, and Morgan Stanley noted Tesla employs more than 135,000 people, over six times SpaceX's workforce. The analysts said this reveals a wide open lane in physical AI that the two companies may be uniquely capable of addressing, separately or, more likely, jointly, though the assessment reflects Morgan Stanley's expectations rather than a confirmed outcome.
IFR reports global humanoid robot sales reached 7,000 units in 2025
The International Federation of Robotics, or IFR, reported that global sales of humanoid robots for industrial and professional service use came to about 7,000 units last year, making it one of the first industry statistical reports to record data on this category of robot. It is the first time the IFR has begun officially collecting data on humanoid robots, after gathering robotics statistics for more than three decades. The figures exclude consumer and military robots, while medical robots are counted in a separate category. IFR Secretary General Susanne Bieller told Reuters that the current number of humanoid robots is still only a tiny fraction of the overall robot population. In 2024, about 542,000 conventional industrial robots were installed worldwide, along with roughly 199,000 commercial service robots for use in transport, hospitality and cleaning. Bieller added that most humanoid robots sold in 2025 have not yet been put to work generating actual output, but were bought by research institutions or technology companies to collect data and train artificial intelligence models. Automakers, the pioneering customer group, are still only testing them in factories in single-digit or low double-digit numbers. Bank of America Global Research estimates that growth will accelerate rapidly, forecasting that global humanoid robot deliveries will jump to 90,000 units this year and could reach 1.2 million units by 2030. The full 2025 statistical report on industrial and service robots will be published officially on September 24.
Toyota to Deploy Up to 400,000 Humanoid Robots at Global Plants From 2028
Toyota Motor Corporation plans to deploy humanoid robots across its global manufacturing operations from 2028, including having them work alongside humans in the same work areas. The Japanese automaker is understood to be planning to gradually deploy up to 400,000 in-house-developed humanoid robots across its global production operations, introduced as part of upgrades of existing factories when new vehicle models are introduced and when new production capacity is added. Toyota is said to be planning to invest JPY 1 trillion, or US$ 6.4 billion, annually from 2028, starting with the introduction of some 150,000 humanoid robots at its headquarters plants, with a further 250,000 planned to be rolled out at the group's 60 production facilities worldwide. The robots to be deployed include Toyota's newly developed in-house ELEY, or Embodied Learning robot for Enhanced Yield, a battery-powered wheeled robot weighing around 50 kg that is said to be capable of tasks such as picking up components from boxes and placing them where required, and folding fabrics. Toyota will centralise the learned data to programme new robots and even create an environment in the future where robots take charge of training new employees, stating that robots are not meant to replace humans, but will coexist with humans.