Tesla Builds First Commercial Optimus Robots at Fremont

Tesla has begun constructing humanoid robots on a new high volume line inside its Fremont, California factory. The company stated in its third quarter shareholder letter that the units coming off this specific assembly line represent the first production quality bots it has manufactured. Tesla added that the current manufacturing process has been proven suitable for a commercial product.

The production space previously built Model S and Model X vehicles until this spring. Engineers cleared the old vehicle line in 46 days to make room for robot production sized for one million units a year. During Wednesday’s earnings call, Elon Musk confirmed that most of the equipment required for that annual capacity is already installed in Fremont. The company expects output to climb toward roughly 20,000 robots per week as the line ramps up.

Musk described the production increase as a classic S curve, noting that the initial slope will be shallow. The engineering team is currently refining the robot’s AI, electromechanical design, reliability, and service life. Musk expects to finish this refinement work by the end of next year. At that point, Tesla plans to start delivering Optimus units to outside companies for dangerous, repetitive, and around the clock jobs.

A central challenge for the new product is data collection. Tesla’s cars send real driving data back to the AI team from millions of vehicles, but Optimus does not yet have a comparable fleet. Musk explained that the lack of a vast fleet of robots limits the available training data. To solve this, the company is creating an Optimus Academy where 10,000 to 20,000 robots will practice a wide range of tasks. Every unit leaving the Fremont line will feed what it learns into the next version of the model.

Hardware upgrades are also scheduled for the coming year. Tesla plans to install its AI5 chip into Optimus next year. Musk said a new Optimus video should arrive within one or two months, showing the robot completing tasks driven by AI rather than preprogrammed motions. He contrasted this approach with rival humanoid demos that rely on scripted routines, arguing that those would not hold up in the real world.