Zoox opened a 220,000-square-foot production facility in Hayward, California, in June 2025. The company said its line could eventually assemble more than 10,000 purpose-built robotaxis annually. At the opening, it was making roughly one vehicle a day, according to reporting from the factory tour.
Those figures describe different things. Ten thousand is design capacity at full operation; one per day was an early production rate. Neither says how many vehicles passed validation, entered a permitted service area, or completed rides. A multiplication of factory capacity by another operator’s rides per vehicle would produce an invented demand forecast.
Zoox’s vehicle has no conventional driver position and seats four passengers facing one another. Building that design itself gives Zoox control over passenger space and vehicle integration. It also means Zoox has to manage manufacturing quality, spare parts, repairs, and the ramp of an unfamiliar vehicle alongside the autonomy system.
The Hayward plant matters because it creates a possible supply path for more than a small pilot fleet. But the case for better unit economics than Waymo’s automotive partnerships is still unproven. A purpose-built cabin might improve the rider experience; it does not automatically reduce cost per paid mile. Vehicle depreciation, utilization, remote assistance, charging, cleaning, maintenance, insurance, and empty mileage remain in the ledger.
Zoox opened public rides in Las Vegas in September 2025 and later listed select-rider service in San Francisco. That is operational progress, though it does not show the Hayward line running at full capacity. The useful milestones now are delivered vehicles, active service vehicles, completed paid rides, and the cost and reliability of operating each car. Zoox has not published enough of that data to calculate a credible economic advantage.