Elon Musk says he wants to show off Tesla’s Optimus robot – just not early enough for rivals to rip it apart frame by frame.

On Tesla’s Q1 2026 earnings call, Musk told investors the company is “a little hesitant” to unveil Optimus V3 too soon because competitors scrutinize every public demo and rush to copy anything that looks remotely useful. In typical Musk fashion, he didn’t name names, but he made it clear that the timing of Optimus’ big reveal is now as much a competitive strategy decision as it is an engineering one. Tesla is targeting a late-summer production start for its humanoid robot at the Fremont factory, with the public unveiling pushed closer to that date than many fans would like.

Underneath the copycat complaint is a huge bet: Musk keeps saying Optimus will be Tesla’s “biggest product ever,” more important than its cars or even its long-promised robotaxis. To lay the groundwork, Tesla is essentially gutting its legacy. The company is sunsetting the Model S and Model X and tearing out their production lines in Fremont so the space can be rebuilt as a robot factory. Internally, Tesla’s guidance has pointed to the Fremont line eventually pushing out up to 1 million Optimus units a year, while a second line at Gigafactory Texas is supposed to scale towards 10 million robots annually later in the decade. For a company that only shipped around two million vehicles last year, those are wild numbers.

Getting there is not as simple as dropping a robot on the end of an existing car line. Musk told investors that dismantling the S/X line in Fremont will take “at least a few months,” and installing a brand-new Optimus line will take “several months” on top of that. If Tesla can go from shutting down one line, ripping it out, installing an entirely new one and turning it on in four months, Musk says that would be “insanely fast” even by Tesla’s standards. That bottleneck – plus his desire to avoid giving rivals extra time to react – is why he’s in no rush to roll Optimus V3 on stage until the hardware and manufacturing are locked in.

The other half of the story is what Optimus actually is in 2026 – and what it isn’t. Tesla’s own updates describe Optimus V3 as a general-purpose humanoid designed to learn skills by watching humans, taking verbal instructions, or even being trained through video demonstrations. The first real-world jobs are deliberately boring: repetitive factory work inside Tesla plants, the kind of tasks that don’t require creativity but demand consistency and stamina. Musk says Optimus will carry “a lot of intelligence” on board so it can work offline, but it will still need direction from what he calls an “orchestration AI” – essentially a manager in the cloud that hands out tasks and priorities.

That is where Grok, the AI chatbot built by Musk’s xAI venture, slides in. Musk has described Grok as the high-level brain coordinating what Optimus does, making sure robots get the right instructions and can be monitored without a human supervisor staring at them 24/7. In his analogy, Grok will talk to Optimus about as often as a manager checks in with their team, meaning the robot should be able to work several hours without oversight before it needs new orders. It is a neat narrative connection between Musk’s AI and robotics bets: Grok handles reasoning and planning, Optimus is the body doing the physical work.

The competitive anxiety Musk is talking about isn’t imaginary. In the last two years, humanoid robotics has gone from sci-fi curiosity to a genuine industrial category. Boston Dynamics has turned its all-electric Atlas from a viral parkour demo machine into a platform that Hyundai is now pushing into automotive production, with its initial 2026 manufacturing run already sold out. Agility Robotics’ Digit is walking around warehouses and logistics facilities, including Amazon sites, lifting and moving totes. Figure AI has signed deals in automotive and retail, Unitree is selling a relatively low-cost G1 humanoid, and companies like Apptronik and 1X are targeting factories and, eventually, homes. Unlike Optimus, some of these robots are already shipping in limited but real commercial deployments.

Musk’s frustration is that every time Tesla shows Optimus doing something new – folding laundry, moving boxes, manipulating tools – the rest of the field pauses the video and treats it like a free blueprint. In a space where nobody has a decade-long head start and where AI software can be replicated far faster than hardware, he clearly believes secrecy is now an asset. Show too much, too early, and you risk giving your next big idea away to rivals with their own factories.

At the same time, Tesla doesn’t have the luxury of staying in stealth forever. Investors who have watched Tesla’s EV growth slow and margins get squeezed by price cuts are looking to Optimus as the next growth engine. Musk has sketched a roadmap where Optimus V3 enters low-volume production this summer, ramps through 2026, and then hits high-volume output around 2027 as the Texas line comes online. Alongside that physical robot, he’s also been talking up a “Digital Optimus” – a joint Tesla–xAI project that uses Grok-style AI to automate knowledge work on computers – as if to say that this brand isn’t just about metal and motors.

There’s also a cultural shift happening inside Tesla. For years, the company’s identity was tied up in being the definitive EV maker: the company that convinced legacy car brands to go electric and then fought them head-on. Now, with Model S and Model X riding into what Musk called an “honorable discharge,” the center of gravity is moving from cars to robots and autonomous systems. The Fremont factory that once symbolized Tesla’s arrival as a real automaker is being quietly reimagined as a humanoid robot plant.

All of that context makes Musk’s “copycat” complaint feel less like a throwaway line and more like a tell. Tesla is no longer the only one playing the game it helped popularize, whether in EVs, autonomy, or now humanoid robotics. Boston Dynamics, Agility, Figure, Unitree, and others are showing that a wave of capable, specialized robots is already hitting the market. Tesla still has advantages – its vertically integrated supply chain, its in-house AI stack, and a global manufacturing footprint – but Optimus has to move from slide decks and stage demos to real, scalable production.

So Musk is threading a needle. He needs to show just enough to keep investors, customers, and talent excited – and to convince the market that Optimus really can be “bigger than cars.” But he also wants to hold back the details that competitors could clone before Tesla has a chance to flood the world with millions of its own robots. That’s why, for now, Optimus V3 is staying mostly behind the curtain, while the factory that will build it gets ripped apart and rebuilt around a bet that humanoids are the next iPhone moment – and that Tesla can get there first.