Thursday October 22, 2026 3:48 pm
Musk Says Optimus Won’t Reach Outside Customers Until Late 2027. The Hands Are the Reason.
Posted by Andru Edwards Categories: Corporate News, Artificial Intelligence

Walking turned out to be the easy part. Asked on Tesla's third-quarter earnings call this week what is still standing between Optimus and a paying customer, Elon Musk went straight to the fingers. "It's mostly the hand," he said.
He put a date on the rest of it, too. Musk said he expects the hand, the reliability and the AI to be worked out "around the end of next year," and that this is "the time to expect when we're delivering Optimus to companies outside of Tesla." Late 2027, then. That is later than the dates Tesla was giving out in 2024.
Fingers are harder than legs
A hand has to do two things that fight each other. It has to be soft enough to pick up something delicate, and tough enough to keep doing that all shift without falling apart. Musk said it is hard to build a hand that "can handle lots of wear and tear" when the fingers are "kind of squishy and have lots of sensors." Squishy and sensor-packed is the opposite of durable, and a robot finger does not heal the way a scraped knuckle does.
His bar for shipping was a robot "that is safe and reliable and has long duration," one that "can't like wear out after a month." Nobody is buying a machine that needs new hands every few weeks, and companies are first in line on Tesla's plan, ahead of any home version.
The AI is not finished either
Musk said there is "still quite a lot of work to do with regard to the AI and the robot." The problem underneath that is data. Tesla has millions of cars feeding driving footage back for Full Self-Driving, and it has nothing comparable for a robot learning to handle objects.
Tesla's answer, as described on the call, is to build roughly 10,000 to 20,000 robots whose job is to practice. Tesla also plans to put its AI5 chip in Optimus around April 2027, which Musk pegged at about ten times the compute of AI4 and called a large jump in capability. He said a new demonstration may land in a month or two, meant to show the robot doing things it was not preprogrammed to do.
What is built, and what is a target
Optimus is in early production on the high-volume Fremont line, the one that used to build the Model S and X, and Musk said "the vast majority of a 1 million unit per line" is already installed there. Tesla's shareholder deck calls the robots coming off it "the first production-quality bots that we have manufactured," and lists California as early ramp and Texas as construction. CFO Vaibhav Taneja said work has started on an Austin factory meant to out-produce Fremont.
The 20,000-robots-a-week figure is a target rate, not an output number. Musk said getting to that annualized rate "takes us probably at least a year, maybe a little longer," and Tesla gave no current weekly count and no cumulative total on the call. His own framing was that "there's like 10,000 things that have to go right for production of the robot," with parts of the design still changing underneath the line.
The dates have moved before
In 2024, the plan was Optimus working inside Tesla in 2025 and external sales as early as 2026. In January 2025, Musk said "the normal internal plan calls for roughly 10,000 Optimus robots to be built this year," allowed that hitting 10,000 by December was "probably not" happening, and said of the ones that did get built, "yes, I'm confident they will do useful things." He later acknowledged that none of them were, per Electrek.
So late 2027 is a plan from a company whose Optimus plans have slipped twice already. That does not make it wrong. It makes it the third date, and Tesla has not met the first two.
Who gets one first
Not you. Musk said the first outside customers would be industrial, handling "dangerous repetitive tasks, things that people really don't want to do," with home units following as production scales. No consumer date was offered.
Musk also said he remains "convinced that Optimus will be the biggest product on earth" and that he would "be stunned if anyone has a robot that is remotely competitive with Optimus." On manufacturing he has a case, because nobody else has a car-plant-scale humanoid line mostly installed. On the robot itself, the schedule now rests on a hand the company says it has not finished designing. The next checkable thing is that demo video in a month or two. Watch what the hands do in it.