Musk says Tesla could have 2 million robotaxis in a year, more than it can build

On Tesla’s Q4/FY2025 earnings call yesterday, Musk mused about the growth rate of its paid robotaxi service, which it launched in Austin this year.

But there’s little evidence that the network is working at the level he claims it is today, and it seems extremely unlikely that it will scale at the rate he claimed either.

On Tesla’s earnings call, a question asked about how well the Robotaxi program is scaling:

What has surprised you about the robotaxi rollout so far, and what factors have constrained fleet expansion to date, which appears to be ~200 vehicles based on public tracking?

Tesla first launched its service in Austin in June, though in quite limited fashion as each car had an operator in the passenger seat. Despite those operators, the cars still had quite a lot of problems, and have suffered a crash rate significantly above that of human drivers.

Advertisement – scroll for more content

But that didn’t stop Musk from promising more. He stated that, by the end of the year, half of the US population would be covered by robotaxis.

Needless to say, that did not happen. Currently, Tesla covers two areas with a service it calls Robotaxi, though its use of that word is suspect in both cases.

In one case, the SF Bay Area, all of Tesla’s ride-hailing vehicles have an operator in the driver’s seat. Tesla is legally barred from offering autonomous taxi rides to the public in California and these rides are mostly available to employees, so the word “Robotaxi” exists solely for branding reasons in this case.

In Austin, the Robotaxi operators have always sat in the passenger seat of the vehicle, until last week when they were banished to a chase car. However, that seemed to just be a temporary stunt, as independent analysis suggested just days later that none of these “driverless” Robotaxis were anywhere to be found.

But in today’s earnings call, that reality was nowhere to be found, and we were instead greeted by some very optimistic words that haven’t really stood up to the evidence available to the public.

Musk answered the above question, stating:

In terms of Robotaxi vehicles carrying paid customers, uh, I think we’re well over 500 at this point between the Bay Area and Austin…. This will probably, y’know, double every month type of thing. It’s gonna, it’s on an exponential curve.

This statement seems suspect, because public evidence doesn’t point to 500 cars being available – and even if they are, combining Bay Area and Austin together doesn’t make a lot of sense, since the systems work differently than one another.

That said, that Musk number lines up with a previous promise of his to have 500 Robotaxis in Austin by the end of the year. Given that Musk is an absentee CEO who spends most of his time trying to advance white supremacy on his twitter feed instead of running his company, maybe he just misremembered his previous promise as if it was the truth (and was probably confirmed by him by the endless self-reinforcing nature of the algorithm he purchased to feed his own ego).

But that’s not the most ridiculous part of Musk’s answer. He also states that the size of the fleet will “double every month,” which seems incredibly unlikely given Tesla’s current production capacity.

If indeed the fleet is currently 500 cars and will double every month, that means 1,000 in a month. Then 2,000 the next. And so on, until the fleet is… over 2 million cars in one year (500 x 2^12 = 2,048,000), by Musk’s estimation.

Now, this may seem like ridiculous extrapolation, but it’s not actually that out of line with statements that Musk has made before. More than once, Musk has claimed that there will be a million robotaxis on the road by the end of whatever-year-he-said-it-in.

Musk has also recently shown that he is unaware of Tesla’s current sales trajectory, suggesting that the company is “making ~2M cars/year and rising,” a number which conveniently lines up with the extrapolation of his prediction of monthly-doubling robotaxi deployments. That number is also ~2!% more than the number of cars Tesla built in 2025.

However, this would require the company to sell zero vehicles (well, he’s certainly working on making that happen), and to instead deploy all of the “~2M” vehicles that it will make over the next year.

Or it would require Tesla cars that are already deployed in the world to end up in the Robotaxi program.

This is something that Tesla has promised before for many years. The story goes that Tesla owners who have FSD could send their cars out as taxis to make money for them. It’s an idea that led Musk to declare that Teslas would be “appreciating assets.”

However, despite promising that for so long, Tesla still has not delivered full self-driving capability to its owners, much less the ability to send cars out to act as autonomous taxis. Tesla has started taking paid rides (in its non-autonomous taxis) on its own, but has not deigned to allow its loyal owners, who paid up to $15,000 up to 9 years ago, to use the software they purchased so long ago.

Rolling this program out to every singe FSD subscriber would get Tesla closer to Musk’s 2-million-by-next-year timeline, but we also learned today for the first time that there are 1.1 million active FSD subscriptions. If every single one of those entered the Robotaxi fleet (and many owners won’t want to do that), that gets us halfway there. This would also likely boost subscriptions, so we could count on that number going up significantly if this was actually rolled out.

But how would that rollout look? Tesla also reiterated today that each geographical area it rolls out to requires a lot of work on corner cases.

Instead of the long-promised infinitely scalable system that could be flipped on fleet-wide with a software update due to Tesla’s massive data advantage, we are now seeing the reality of individual ground truthing being required in every new bit of city the Robotaxi fleet expands to.

Scaling that to an area that could be served by 2 million vehicles seems daunting – and it certainly won’t be a release that “doubles every month.”

And so, this seems like another example of Musk just talking without considering the implications of what he’s saying, as we’ve seen plenty of before. How long is anyone going to keep accepting this nonsense at face value?

Charge your electric vehicle at home using rooftop solar panels. Find a reliable and competitively priced solar installer near you on EnergySage, for free. They have pre-vetted installers competing for your business, ensuring high-quality solutions and 20-30% savings. It’s free, with no sales calls until you choose an installer. Compare personalized solar quotes online and receive guidance from unbiased Energy Advisers. Get started here. – ad*