Tesla (Nasdaq: TSLA): Optimus and the Hardest Manufacturing Ramp Tesla Has Attempted
Tesla shut down the Model S and Model X line at its original Fremont factory in 2026 and converted the space to build Optimus, its humanoid robot, at a design capacity of one million units a year. As of its most recent earnings call, Tesla had disclosed zero verified production units — and Elon Musk himself has spent multiple quarters actively lowering expectations for how fast this ramp will go.
Who they are
Tesla's humanoid robot, Optimus, was first announced in concept form at the company's AI Day in August 2021 and has moved through several public prototype demonstrations since. The genuinely significant 2026 development is physical, not conceptual: Tesla ended production of the Model S and Model X in the second quarter of 2026 and converted that dedicated assembly space at its original Fremont, California factory into an Optimus production line, rather than retooling for another vehicle.
That conversion is a real, visible capital commitment — Musk posted a photo of himself with the Optimus production team standing inside the converted line, and Tesla's own Q2 2026 shareholder deck confirmed the Model S/X teardown was complete, with Optimus lines being installed and production "anticipated later this year." The stated eventual design capacity for the Fremont line is one million Optimus units annually, a figure Tesla and outside analysts have both been careful to describe explicitly as a long-term design target rather than a near-term production forecast.
What they do
As of the Q2 2026 earnings call on July 22, 2026, Tesla had not disclosed a single verified Optimus production unit from the new Fremont line. The company's own Q2 shareholder update described first-generation Optimus lines as "being installed" at Fremont with production "anticipated later this year" — language notably more cautious than earlier-year statements. On the earnings call itself, Musk described Optimus as "the hardest product to scale manufacturing that we've ever made," citing the newness of the approximately 10,000 unique components in the robot and the absence of an existing supply chain to draw from, and explicitly said the initial portion of the manufacturing S-curve "will be quite flat and long."
Musk has also been explicit that public demonstrations circulating online of humanoid robots performing dexterous tasks are, in his words, "pre-programmed or remote controlled" rather than genuinely autonomous, and that no humanoid robot yet exists that can perform generalized tasks the way Optimus is intended to eventually. That is a notably candid admission from Tesla's own CEO about the state of the industry broadly, not merely a caveat about his own product — and it directly contradicts some of the more optimistic framing found elsewhere in humanoid robotics coverage.
Multiple third-party production estimates for 2025 put cumulative Optimus builds at only a few hundred units, all currently used internally for what Musk has described as "learning," not productive commercial deployment. No verified third-party customer relationship or external deployment has been confirmed as of this writing.
How they make money
Tesla's overall business remains almost entirely automotive, energy storage and services — Optimus generates no disclosed revenue whatsoever as of Q2 2026. Total Q2 2026 revenue was a record $28.24 billion, up 26% year over year, but profitability compressed sharply: operating margin fell to 1.4% from 4.1% a year earlier, automotive gross margin excluding regulatory credits was 16.3%, and the company posted negative free cash flow of $1.09 billion despite the record top line, driven by capital expenditure of $5.79 billion, up 142% year over year, tied to AI, robotics and manufacturing investment.
Energy storage is the one segment showing clear strength alongside the Optimus buildout, with Q2 2026 deployments of 13.5 GWh, up more than 40% year over year, and energy revenue of $3.14 billion — a reminder that Tesla's near-term financial health depends far more on cars and batteries than on any humanoid robot timeline.
Where it sits in the value chain
Tesla sits in Stage 7 — Robot builders, and its approach is the most vertically integrated of any name in this chain: Musk has stated Tesla intends to manufacture its own planetary roller screws in-house rather than purchasing them from specialist suppliers at what he's called aerospace pricing, and the company is separately developing its own AI inference chip for Optimus rather than adopting a platform like NVIDIA's Jetson Thor or Qualcomm's Dragonwing.
That vertical integration means Tesla appears less frequently as a named customer in other companies' investor disclosures than one might expect given Optimus's public profile — the widely reported $685 million Sanhua actuator order was denied by Sanhua itself, and most confirmed component relationships in the Chinese supply chain, such as Leaderdrive's reported small-batch R&D orders, remain sample-scale rather than confirmed production commitments.
The bigger trend
Tesla's Optimus program is simultaneously the highest-profile and, on current disclosed evidence, one of the least advanced humanoid programs relative to its public attention: Unitree and AgiBot are already shipping thousands of units in China, UBTech has begun mass production and delivery of its Walker S2 with disclosed orders exceeding RMB 800 million, and Boston Dynamics has committed all of its 2026 Atlas production to internal and partner use. Tesla, by contrast, has yet to disclose a single verified production unit from its new dedicated line.
That gap between attention and disclosed progress is not necessarily a sign Tesla will fail — the company has a long history of slow initial ramps eventually reaching massive scale, most obviously with Model 3 production a decade earlier. But it is a gap worth being precise about: as of mid-2026, Optimus remains a capital commitment and an engineering effort, not yet a demonstrated manufacturing or commercial success, by Tesla's own management's account.
What to watch
- The first disclosed verified Optimus production unit count from the Fremont line, expected "later this year" per Tesla's own Q2 2026 guidance.
- Any confirmed external customer or deployment announcement, distinct from internal "learning" use.
- Whether Tesla's in-house planetary roller screw and AI chip manufacturing strategies achieve cost targets closer to the stated $20,000 long-term robot price goal.
- Cortex supercomputer capacity progress at Giga Texas, the compute infrastructure underlying Optimus's AI training.
- Whether the "flat and long" S-curve Musk described begins to show visible acceleration in subsequent quarterly disclosures.
Related companies
🇨🇳Already shipping and delivering humanoid units at a scale Tesla has not yet disclosed — not yet profiled.
🇨🇳Reported (and denied) as a Tesla actuator supplier, illustrating the murkiness around Tesla's actual supply chain.
🇰🇷A smaller-scale humanoid developer pursuing a platform strategy rather than Tesla's full vertical integration.
FAQ
How many Optimus robots has Tesla actually produced?
As of Tesla's Q2 2026 earnings call in July 2026, the company had not disclosed a single verified production unit from its new Fremont production line. Third-party estimates for cumulative builds through 2025 put the figure at a few hundred units, all used internally for training and development rather than commercial deployment.
Does Tesla have any confirmed Optimus customers?
No external commercial customer has been confirmed by Tesla as of this writing. Reports of large component orders, such as a widely circulated $685 million actuator order with Sanhua Intelligent Controls, have been denied by the supplier named in the reports.
What did Elon Musk say about Optimus's manufacturing ramp?
On Tesla's Q2 2026 earnings call, Musk described Optimus as "the hardest product to scale manufacturing that we've ever made," citing roughly 10,000 unique components and no existing supply chain, and said the initial phase of the manufacturing ramp "will be quite flat and long." He also said no humanoid robot industry-wide, including Optimus, can yet perform genuinely generalized tasks outside pre-programmed or remote-controlled demonstrations.
What is Tesla's target price and production scale for Optimus?
Musk has referenced a long-term target price around $20,000 per unit and a Fremont factory design capacity of one million units per year, but both figures are long-term aspirational targets rather than near-term commitments, and Tesla has not disclosed a timeline for reaching either.