Tesla's Optimus hands slow the climb from hundreds of robots a week to 1,000 by year-end

At the year-end target of 1,000 a week, Tesla would be building Optimus at about 52,000 robots a year ^[1]. The longer-run goal of 20,000 a week is 20 times that, a little over a million a year ^[2]. Output already runs in the several hundreds, so the multiple left to cover before December is smaller than the tenfold Tesla has just made ^[3]. By The Information's account, the hand is one of the biggest obstacles in what is left ^[2].

The hand needs several mechanical elements to work together precisely enough to grasp, carry and manipulate objects, and it has to hold up through many rounds of industrial work ^[7]. Durability is also what keeps the V3 (the lighter version with better cameras now coming out of Fremont ^[4]) from being Tesla's final commercial robot. It has to improve before the robot is released in large numbers ^[8]. A year-end rate of 1,000 a week would be 1,000 a week of an interim design, doing tightly defined tasks in supervised areas in Palo Alto and on factory floors ^[6].

The sales plan is built around that. A lease lets Tesla take a robot back, replace mechanical components and send it out again, and data from the customer's floor goes into training the AI ^[10]. Tesla has not started selling Optimus to outside buyers ^[14], and the reporting does not include lease pricing.

If Tesla reaches 1,000 by December with the hand line fixed, the open question moves to how the hands hold up in leased fleets. If suppliers and hand assembly keep the ramp in the hundreds, the year-end target slides into 2027. A third case sits between them. The line hits the rate, but robots come back for parts often enough that the returns, and not the factory, set how many are working at any one time.

Jerry Wang, global executive chairman of Faraday Future, a former electric-vehicle maker that has moved into robotics, sees the limit elsewhere. "The hardware is not a barrier right now. The barrier is software and training. This will take time, but I think that they will get there," he said ^[11]. Asked about a 2027 commercial launch, he said: "I think it's very realistic" ^[12]. His company says it has delivered more than 500 robots this year across humanoid, four-legged and wheeled types ^[13].

I think 1,000 a week is the right test of the factory and an incomplete one for the stock. A leased interim robot that keeps coming back for new hands produces little value however fast it is built. The counter-case is that the lease fleet is a durability test Tesla controls, and a V3 that stays out working would shorten the path to the commercial robot. JPMorgan said in a note last month that commercial sales of Optimus Gen 3 may begin in the second half of 2027 ^[15]. This view is wrong if Tesla reaches 1,000 a week by December and the leased robots stay in the warehouses, because then the hands were a production problem Tesla has already solved.