Post

Tesla Still Has No Weekly Optimus Rate. The Hands Are Still Hand-Built.

A gloved hand uses a screwdriver on a white robot hand and forearm on a factory fixture, with the headline “Optimus Delay, No Weekly Rate” and a red bar reading “Hands Still Hand-Built.”

Tesla still has not published a weekly Optimus production number. The mid-2026 V3 reveal did not happen. The first official builds are for Tesla’s own training program, not a shop page.

That is the delay.

Friday’s report from The Information is not a Tesla IR release. It is people familiar with the project describing a line that is moving — and a robot that is still late where it counts. The hands and forearms on the V3 units coming off that line still have more than 100 screws and small parts a person has to assemble. Fixtures miss. New robots need rework. The AI still needs task-specific programming in controlled areas.

Tesla already told you this ramp would be slow. On July 1, the same day he posted a photo walking the Fremont Optimus line, Elon Musk wrote in a reply on X:

“No, Optimus production will be extremely slow at first, as everything is new. This is not like making a car.”

That is the story.

Tesla already said the first stretch would be flat

On January 20, 2026, Musk posted the S-curve warning himself. Initial production is always very slow. “For Cybercab and Optimus, almost everything is new, so the early production rate will be agonizingly slow, but eventually end up being insanely fast.”

Two days later at Davos he put a customer date on it that was not 2026. Public sales, he said, come “by the end of next year” — end of 2027 — and only when reliability, safety, and the range of functionality are “very high.” Axios, WSJ, and Fox Business carried that language.

The missed dates sit behind that.

The dates Tesla missed

January 29, 2025. Q4 2024 earnings call. Fortune reported the next day that Musk described the internal plan as “roughly 10,000 Optimus robots” in 2025 and said several thousand would be “doing useful things” by year-end. That useful-work target did not land. He did not lock 10,000 as a promise.

January 28, 2026. Q4 2025 earnings call. Musk walked the factory claim back. The call transcript and coverage in Barron’s and PYMNTS carry the same language: Optimus was “still in the R&D phase.” It had done “some basic tasks.” It was “not in usage in our factories in a material way.” Significant volume, he said, would wait until “probably the end of this year.”

Same call: he said Tesla would replace the Fremont Model S and Model X line “with a 1 million unit per year line of Optimus.” CNBC carried the quote. That number is a long-term design target for that floor. It is not a 2026 production rate.

March 31, 2026. The Q1 V3 reveal Tesla had pointed at did not happen. Musk said the robot was walking around but still needed “finishing touches.”

April 22, 2026. Q1 call. Production start moved to late July or August. The V3 reveal was “probably middle of this year,” held back because competitors copy designs “frame by frame.” On the year’s output he said: “I don’t know what the production rate of Optimus will be this year. It is impossible to predict these things.” About 10,000 unique parts. “Quite slow.”

July 1, 2026. Musk posted a photo walking the Fremont Optimus line. The same day, replying to a fan theory, he wrote that production would be “extremely slow at first” because everything is new. “This is not like making a car.”

The last official word: July 22

The Q2 2026 shareholder update is the last official production language Tesla published:

“We have decommissioned the manufacturing lines for Models S & X at our Fremont Factory and are installing the first-generation lines for Optimus, where we expect to start production soon.”

“The initial Optimus builds will be used in our Optimus Academy for training data collection and further functionality development.”

Same deck. California Optimus and Texas Optimus are both listed as Construction. Tesla’s own footnote on that table: installed capacity is not current rate. The S and X floor is how Tesla made room. It is not why the robot is late.

On the July 22 call he went further. Optimus, he said, will be “the hardest product to scale manufacturing that we’ve ever made at Tesla because everything on the robot is new.” With cars, a lot of the non-electric parts already had a supply chain. “With Optimus, there is no supply chain.” The ramp follows an S-curve whose first stretch is “quite flat and long.”

Tesla has not published a weekly Optimus unit count in a deliveries release, a 10-Q, or that Q2 deck. Late September. Still no V3 event. Still no tesla.com Optimus listing.

What The Information reported Friday

The new story is not Tesla’s. It is The Information, dated September 25, by Qianer Liu and Grace Kay, citing people familiar with the project.

Those sources said Tesla built several hundred Optimus robots a week in August, up from a few dozen a week in the second-quarter pilot. That is the “tenfold” figure now circulating. It is a source claim. It is not a Tesla IR number.

The same people said managers told staff they want a continuous automated line above 1,000 robots a week by year-end. The Information also used “around 20,000 a week” as the eventual target. That last figure is roughly the weekly math on Tesla’s already-stated 1 million-a-year Fremont design. Tesla has not confirmed 1,000 a week as a 2026 exit rate.

The snags in that report cluster on the hands.

The Information said each V3 hand and forearm has more than 100 screws and other small parts that still have to be assembled by a person. Robot parts are smaller than car parts and have to fit tighter. Fixtures at the hand, joint, and electronics stations do not always line the parts up. New robots come off the line needing fixes. Some touch sensors have been unreliable enough that Tesla is developing a replaceable “sensing glove” so a failed sensor does not take the whole hand with it.

The same report said Optimus still needs task-specific programming in controlled areas, not general-purpose work, and that some factory workers pushed back on wearing motion-capture gear to train a robot designed to replace them. Tesla, those sources said, moved that data collection onto dedicated teams.

The Information also said the V3 units on the line are not the customer version, and that early commercial units may be leased so Tesla can pull them back and update them.

Treat every sentence in this section as reported, not Tesla-certified. Tesla has not confirmed the weekly rate, the screw count, the glove, the worker-suit story, or the lease plan.

What this does not mean

Friday’s report and Tesla’s own record rule out six easy readings:

  • It does not mean Tesla is building 20,000 Optimus robots a week. That is The Information’s reported eventual target, roughly the long-term Fremont design converted to a weekly number. Capacity is not current rate.
  • It does not mean Tesla published a 2026 weekly run-rate. It did not.
  • It does not mean you can order one. There is no tesla.com Optimus shop listing as of this writing.
  • It does not mean the first Fremont units are customer robots. Tesla’s own deck sent those builds to Optimus Academy.
  • It does not mean the AI is “solved.” On the Q2 call Musk described the unsolved bar as a robot that can do a task if you ask it or show it a video, with no programming, and Tesla has not said Optimus cleared it.
  • It does not mean Model S and Model X owners lose service. Tesla said it would keep supporting the cars, and that is a separate story.

If you own a Tesla

If you bought the “Tesla is a robotics company” thesis with a Model 3, Model Y, or Cybertruck: the official record is still construction, a slow first S-curve, and training-data robots. The Information’s August “hundreds a week” figure, if it holds, is still a long way from the million-a-year line Tesla described in January. Tesla has not put that figure on a slide.

Anyone waiting to buy an Optimus: there is no order page, no published price, and no Tesla-issued weekly rate. The Information says early units may be leased to a short list of companies so Tesla can pull them back. That is a source claim. Wait for Tesla to say it.

The delay is not a mystery Tesla hid. Musk already said everything is new and the first stretch is flat. Friday’s report is the part he did not put on a slide: the hands are still built by people.

Answers

Is Optimus delayed?

Yes, against Tesla’s own earlier dates. Several thousand doing useful work by the end of 2025 did not happen. The mid-2026 V3 reveal did not happen. Tesla has not published a weekly rate. First official builds go to Optimus Academy, not customers.

Is Tesla mass-producing Optimus?

Tesla says first-generation lines are going in at Fremont and that production was expected in 2026. Tesla has not published a weekly count. The Information’s sources put August in the hundreds per week.

How many Optimus robots is Tesla building?

Tesla has not said. Do not treat a reporter’s “several hundred a week” as a Tesla number.

Can I buy one?

Not from Tesla’s shop today.

Is the 1 million a year target current output?

No. That is the long-term design target Musk stated for the Fremont floor. Installed capacity is not current rate.

Why the delay?

Tesla’s version: new product, no existing supply chain, first part of the S-curve is long and flat. The Information’s version: hands still hand-built, fixtures miss, AI still task-specific.

Sources

Primary sources and the reporting this story relies on:

  • Elon Musk on X, January 20, 2026 and July 1, 2026
  • Tesla Q2 2026 Shareholder Update and earnings call, July 22, 2026
  • Tesla Q1 2026 earnings call, April 22, 2026
  • Tesla Q4 2025 earnings call, January 28, 2026 (transcript, PYMNTS, CNBC)
  • Fortune, January 30, 2025 (Q4 2024 call quotes)
  • World Economic Forum remarks at Davos, January 22, 2026 (Axios, Fox Business)
  • The Information, Qianer Liu and Grace Kay, September 25, 2026 (attributed, not a Tesla disclosure)

Also in The Brief: Tesla Semi Deliveries Just Started. 500 Miles Loaded. $260K–$290K.

Come see tech in a new light.

The RED Review · theredfire.com · YouTube @TheREDReviewTech · X @TheREDReview_EV