Tesla Optimus Is Finally in Production. Here Is What That Does — and Doesn't — Mean

Tesla has been showing us Optimus since 2021. In its Q3 2026 update it finally said something concrete: the robots now coming off its new Fremont line are, in Tesla's words, "production-quality", the first it has ever built to that standard, and the process is ready for a commercial product.

That's a real milestone. It's also very easy to read too much into, so I want to go through what Tesla actually said and what it didn't.

What actually happened

Tesla tore out the Model S and Model X line at Fremont, which had been building cars for 14 years, and cleared the space in 46 days. Most of the equipment for a line designed to make about a million robots a year is now in place there.

Production has started, but slowly. On the earnings call Musk called it "early production" on the high-volume line.

The robot coming off that line is Optimus Gen 3, the first version designed from scratch to be mass-manufactured. Its hands have 22 degrees of freedom. Tesla plans to switch to a Gen 4 design by the time its second Optimus factory, in Texas, comes online.

How far there is to go

Musk's own numbers tell the story better than anything I could add.

Musk said something like 10,000 things have to go right and described the early ramp as shallow and bumpy. He's also admitted to problems with the hands and with getting rare-earth magnets out of China. Neither is a small issue. Hands are generally seen as the hardest part of making a humanoid useful, and there's a magnet in every motor.

Why I'm still skeptical

Tesla said it would build 10,000 Optimus units in 2025. It didn't. For 2026 it hasn't given a volume target at all, and Optimus still doesn't get its own line in Tesla's quarterly production numbers. Zacks analysts argued earlier in the year that the story was running ahead of the evidence, and the Q3 update doesn't include an independent production count that would change that.

My read is this. Tesla has shown it can build Optimus to a production standard. It hasn't shown that it can build a lot of them, or that they do useful work for less than it costs to pay a person or buy a conventional robot arm.

Where Optimus sits against Atlas and Digit

Tesla isn't first. Boston Dynamics began building its commercial electric Atlas in January, and every one made this year is spoken for by Hyundai and Google DeepMind. Agility's Digit has been doing paid warehouse work for GXO since 2024 and has more than 65,000 hours on the clock.

What Tesla has that neither of them can match is factory scale and money. Its capital spending guidance for 2026 is over $25 billion, and the Optimus lines are part of that. If the ramp works, Tesla could out-produce the rest of the Western humanoid industry combined. That's a big if.

What would change my mind

1. The demo video. Musk said one might come within a month or two of the earnings call. I want unedited footage of Optimus working on its own, not something teleoperated.
2. An actual production number. Even a small weekly figure would tell us more than every past target put together.
3. The first outside contract. The day a company that isn't Tesla signs up to use Optimus is the day it becomes a business.

You can see how Optimus stacks up against the other humanoids on our comparison page.