Tesla is building hundreds of Optimus robots a week, but leasing them only to factory-like sites. Here's what that means for contractors
Tesla reportedly lifted Optimus output from a few dozen to several hundred a week, yet the robots are still hand-limited and will be leased to a short list of customers with facilities like Tesla's. For contractors, the near-term opening is the prefab shop and laydown yard, not the open jobsite.
Tesla's Optimus humanoid robot is moving from prototype to volume, but it is not coming to your jobsite soon. Reporting this week says Tesla's output jumped from a few dozen robots a week to several hundred, while the company plans to lease the first units only to customers whose facilities look like its own. For contractors, that points to the prefab shop, fabrication bay, and warehouse before it points to the field.
What was actually reported?
The figures come from reporting by The Information, relayed by Electrek, The Next Web, and others. The key points:
- Output rose from a few dozen robots a week in Q2 to several hundred a week in August.
- Managers are reportedly aiming for an automated line making more than 1,000 a week by year-end, with an eventual target around 20,000 a week.
- Each hand and forearm has more than 100 small parts that workers still assemble by hand, which is the main bottleneck.
- Most units are used inside Tesla for testing, training, and data collection. The current V3 robots are reportedly not the version Tesla plans to commercialize.
- Tesla reportedly intends to lease, not sell, to a short list of companies with factories and warehouses similar to its own, then use that deployment data to improve the AI.
These are press reports of internal targets, not Tesla commitments. Treat the 1,000-a-week and 20,000-a-week numbers as aspirations.
Why does the lease-first plan matter to a contractor?
The lease detail tells you more than the production count. Tesla is choosing sites where the floor is flat, the layout is stable, and tasks repeat. Reports say the AI still depends on task-specific training and can behave unpredictably in unfamiliar situations.
A construction site fails each of those tests. Conditions change daily, trades overlap, and the "layout" is whatever the last crew left behind. A robot that needs a familiar environment is a poor fit for a building under construction.
Where could humanoids show up first in construction?
| Setting | Fit with what's been reported | Example tasks |
|---|---|---|
| Prefab / modular shop | Strong: fixed layout, repeat tasks | Moving parts between stations, kitting, loading racks |
| Fabrication bay (duct, pipe, rebar, steel) | Moderate | Material staging, part sorting |
| Warehouse / laydown yard | Moderate | Pallet moves, inventory staging |
| Active jobsite | Weak for now | Unstructured, changing, safety-critical |
What should a PM or ops director do now?
- Don't budget for it. No pricing, lease terms, or availability for outside customers has been reported. Anything you plan around now is speculation.
- Ask what your prefab and warehouse tasks look like. If you run a shop with repeatable material handling, write down the tasks and their cycle times. That is the data a lease conversation would start with.
- Expect lease terms to carry data clauses. Tesla reportedly wants deployment data and the right to retrieve units for upgrades. Before any robot works around your crews, check who owns the footage and logs, and how incidents are reported.
- Check your insurer and safety plan. A robot working near people in a shop changes your risk profile. Ask your carrier before a pilot, not after.
- Watch the hands. The hand-assembly bottleneck is the clearest sign of how fast this scales. If reports show it solved, the timeline moves up.
The takeaway
Humanoid robots are getting built in volume, but the evidence so far says they will start in controlled, repetitive sites. If you run a prefab operation or a large warehouse, start documenting your repeatable tasks now. If you run mostly field work, a robot on your jobsite is still a long way off, and this week's news does not change that.
- Can Tesla Optimus robots work on a construction site?
- Not yet. Reports say Optimus is being leased first to companies whose factories and warehouses resemble Tesla's, because the robots struggle with unfamiliar settings. A changing, unstructured jobsite is the opposite of that.
- How many Tesla Optimus robots is Tesla making?
- According to The Information's reporting as relayed by several outlets, output rose from a few dozen a week in Q2 to several hundred a week in August. Managers are reportedly targeting more than 1,000 a week by the end of 2026.
- Will contractors be able to buy an Optimus robot?
- Reportedly not at first. Tesla is said to plan leasing the robots to a short list of customers rather than selling them, partly so it can collect deployment data and pull units back for upgrades.
- What is holding Optimus back?
- Reports point to hand assembly, with more than 100 small parts per hand and forearm still fitted by workers, plus touch-sensor reliability and AI that depends on task-specific training.