Tesla Targets 10 Million Optimus Units at Texas Plant

Tesla aims to produce 10 million Optimus robots a year in Texas.

Tesla signaled a robotics first strategy on its Q1 2026 earnings call, bolstered by $3.9 billion in operating cash flow and a 21% GAAP gross margin. Production data shows the company intends to shift aggressively into humanoid automation, laying groundwork for a high-volume push that would dwarf current automation scales in most manufacturing sectors.

In Fremont, California, Tesla plans to begin Optimus production in the second quarter, with the first-generation line designed to output one million units annually. The company will replace the existing Model S and Model X production lines to make room for this first-generation robotics plant, a strategic pivot that makes clear the seriousness behind the robot program. At the same time, site preparation is already underway for a second-generation line at Giga Texas, envisioned with a long-term capacity of 10 million robots per year. The Texas project is framed as a step toward industrial-scale humanoid manufacturing, with a dedicated effort to build out a self-contained robotic production ecosystem.

Tesla has framed the plan as more than a demo or pilot. The initiative is supported by a broader push into vertical integration, including development of the AI5 inference processor and what the company calls a “Digital Optimus” intelligence layer. In Fremont, the first-gen line is engineered for a defined, smaller-scale ramp, while the Texas campus is being designed to absorb a far larger, more ambitious production cadence. Floor space, power, and the integration infrastructure needed to sustain tens of thousands of robots operating in parallel remain crucial questions that Tesla has not publicly disclosed for either site.

From an industry perspective, the shift highlights two hard truths about scaling humanoid automation: throughput is a function of more than robot count, and the real bottlenecks often lie in integration, data, and maintenance ecosystems rather than the robots themselves. Integration teams report that turning a factory into a robotics-first cell means redesigning tool cribs, physical layouts, and material-handling logic to avoid cycles where a robot waits for a task or a piece of tooling. The result is not just more robots, but a new, highly integrated digital backbone that coordinates perception, planning, and actuation across a multiyear ramp.

There are tangible caveats that any plant manager should watch for. First, the exact floor space and electrical load required to sustain a 10 million-unit per year line have not been disclosed, and those figures will drive capex, zoning, and safety considerations for years to come. Second, even with a 10X scale in sight, human workers will still be essential for programming, debugging, and unexpected maintenance, especially in the early phases as software and hardware converge. Finally, hidden costs—safety systems, spare-parts inventories, software updates, and ongoing training—will nibble away at initial ROI assumptions until the system matures.

If Tesla can translate this ambitious blueprint into reliable uptime and predictable cycle times, the payback calculus could be compelling for industrial automation at large. The company’s numbers suggest a transition from one million units a year in California to a long-term, ten-million-unit cadence in Texas, a leap that would redefine what a factory can be when robots, chips, and software are co-located at scale.

Sources

From EVs to robotics: Tesla targets 10M Optimus units with new Texas plant