"content": "Good afternoon, everyone, and welcome to Tesla's Fourth Quarter 2025 Q&A Webcast. My name is Travis Axelrod, Head of Investor Relations, and I'm joined today by Elon Musk, Vaibhav Taneja, and a number of other executives.\n\nOur Q4 results were announced at about 3:00 p.m.

Central Time in the update deck we published at the same link as this webcast. During this call, we will discuss our business outlook and make forward-looking statements.

These comments are based on our predictions and expectations as of today. Actual events or results could differ materially due to a number of risks and uncertainties, including those mentioned in our most recent filings with the SEC.\n\nDuring the question-and-answer portion of today's call, please limit yourself to one question and one follow-up.

Please use the raise hand button to join the question queue. Before we jump into Q&A, Elon has some opening remarks.

Elon?\n\nThanks,

Travis. We've updated the Tesla mission to amazing abundance, and this is this is intended to send a message of optimism about the future.

I think we're most likely headed to an exciting amazing era of abundance. I think with the with the advent or with the continued growth of AI and robotics, I I think we actually are headed to a future of universal high income, not universal basic income, but universal high income.\n\nThere there's going to be a lot of change along the way, but that is that is the that is what I see as the most likely outcome.

I think that's that that it makes sense to update Tesla's mission to reflect that that goal.\n\nAnd and obviously along that way, we're going to keep improving safety, driving down the cost of goods, and and giving people access to anything they need without compromise and still making sure that the environment is great, nature is great, and and people can have whatever they want, which seems like probably the best future. I'm open to other ideas, but that sounds like that sounds like that sounds like if you could say what is the best future you could possibly imagine, I I guess it would be that everyone can have whatever they want, including amazing medical care, and and but we still keep, the beauty of nature and and and Earth.\n\nI think that's probably the best outcome.

We're we're seeing obviously the first steps along that way this year for Tesla, first major steps, as we increase vehicle autonomy and begin to produce Optimus robots at scale.\n\nWe're we're making very very big investments, this is going to be a very big CapEx year, as Vaibhav will will get into. That that is deliberate because we're making big investments for an epic future.

I think all these investments make a lot of sense. We'll continue to make sure that when we do spend capital, it is spent very efficiently.\n\nBut but it's it's a lot of things.

Major investments in in in batteries and the entire supply chain for batteries. We're also going to be significant manufacturers of of solar cells, and and we're making massive investments in AI chips.\n\nSo, but I think these all make a ton of ton of strategic sense.

Then I guess I have like one I guess it's it's not exactly bad news, but it's it's a it it's it's what we're we're It's It's time to basically bring the Model S and X programs to to an end with an honorable discharge, because we're really moving into a future that is based on autonomy. If if you're interested in buying a Model S and X, now would be the time to order it, because we we expect to wind down S and X production next quarter and and basically stop production of Model S and X next quarter.\n\nWe'll obviously continue to support the Model S and X programs for as long as people have the vehicles, but but we're we're going to take the Model S and X production space in our Fremont factory and convert that into an Optimus factory, which will with the long-term goal of having a million units a year of of Optimus robots in the current SX space in in Fremont.

That is slightly sad, but but it's but it is it is it is time to bring the SX programs to to an end and shift shift really It's It's part of our overall shift to an autonomous future. As my profile picture on on X said for a few months there, the future is autonomous.\n\nAnd so let's let's say with respect to Full Self-Driving and Cybercab, I think people are obviously following with very close attention the progress of FSD.

You can experience it for yourself if you if you've got a Tesla. You can You notice with really with with every software update, the car gets better and better at autonomy.

We're, we we we were able to do our first rides with no safety monitor in the car in Austin. These were paid rides, these were I just sort of randomly selected paid rides with no safety monitor.

We're I think maybe as of maybe yesterday or so, we we actually don't We don't even have a chase car or anything like that. These are just cars with no people in them, and no one's following the car in in Austin.

We're we obviously are being very cautious about this because we want to have We want We want to have no no no injuries or or serious accidents along the way. I think it makes sense to be very cautious, but you'll see the amount of autonomy increase dramatically, I think, every every month essentially.\n\nSo, and and then and and then there'll also be an opportunity, something we've talked about for a long time, for existing owners of of Teslas to add or subtract their cars to the fleet, kind of like how Airbnb works, where you can add or subtract your house to the the Airbnb inventory.

I think probably the value of the the Tesla's this sort of part people adding or subtracting their cars to the Tesla autonomous fleet is probably a little underweighted, underweighted by a lot of people, because we've got millions of cars with AI4 that can do this.\n\nSo, so that that it might potentially I think it will provide an opportunity for a lot of customers to earn more by lending their car to the fleet than their lease cost to Tesla. Which is kind of it's kind of like you you get in that scenario, you basically get paid to own a Tesla.

It's quite a good scenario.\n\nAnd and we we expect to have fully autonomous vehicles in probably, I don't know, somewhere between a quarter and half of the United States by the end of the year, pending regulatory approval. A big factor would be if if there's some kind of federal preemption for autonomous vehicles.

In the absence of that, you kind of have to go in a city-by-city or state-by-state basis. Nonetheless, even if it is city-by-city, state-by-state, we we expect to be in, I don't know, dozens of cities, dozens of major cities by the end of the year.\n\nWith respect to energy, the the Tesla Energy team is done incredible work, and the growth rate on that is continuing to be very strong.

We're we're building more manufacturing capacity and and expect that, yeah, and energy will will have very high growth for really as far into the future as we can imagine. The The solar The solar opportunity is underestimated.

We think the the best way to add significant capability to the grid is or energy to the grid, if we're let's say powering AI data centers, is solar and batteries on Earth and solar in space. That's why we're we're going to work towards getting 100 GW a year of of solar cell production, integrating across the entire supply chain from raw materials all the way to finished solar panels.\n\nMaybe a bit more about Optimus.

We'll probably unveil Optimus 3 in a few months. I think it's it's going to be quite surprising to people.

It's a an incredibly capable robot. As I mentioned, we are replacing the SX line in Fremont with a million-unit-per-year line of Optimus.

Now because it is a completely new supply chain, it's just it's a There's really nothing from from the existing supply chain that exists in in in Optimus. Everything is designed from physics first principles.\n\nSo that means the the normal S-curve of manufacturing ramp will be longer for Optimus than it is for products that have at least some portion of an existing supply chain.

Like when everything's new, the the production rate will be proportionate to the least lucky, least confident part of the entire supply chain. If there's 10,000 things that need to go right, it's, it only takes one to to be slow to to lag that.

It will be sort of a stretched-out S-curve, but I'm I'm confident that we'll get to a million units a year of in Fremont of Optimus 3.\n\nAnd and this this Optimus 3 really will be a a general-purpose robot that can learn by observing human human behavior. You can like demonstrate a task, or literally verbally describe a task, or show it a task, even show it a video, and it will be able to do that task.

It's a it's going to be a very capable robot. I I think long-term Optimus will have a very significant impact on the US GDP, like it will actually move the needle on US GDP significantly.\n\nSo, in conclusion, there's still obviously many who doubt our ambitions for creating amazing abundance, but we're we're confident it can be done and we're that we're making the right moves technologically to ensure that it does.

Tesla's obviously not It never been a company to shy away from solving some of the hardest problems. That it's I think that's kind of how you build value in a company, is you solve hard problems.

It's like I don't know how you create value by solving easy problems.\n\nSo, there's there's a lot of hard problems that the Tesla team is going to solve, but it's an incredibly talented, hardworking team. I'd like to thank actually everyone at at Tesla for their their incredible hard work, and it's an honor to work with such a talented group.

Thank you to everyone who is supporting this mission. The future is more exciting than you can imagine.\n\nFantastic.

Thank you so much, Elon. Next we have some remarks from Vaibhav.

Go ahead.\n\nThanks,

Travis. Q4 '25 was an interesting quarter in a couple of respects.

On the autos front, while in Q3 we saw a surge in US demand before the IRA consumer credit cliff, pulling in some demand from Q4. In other parts of the world, we saw increase in demand, leading to record deliveries in smaller countries like Malaysia, Norway, Poland, Saudi Arabia, and Taiwan, while continued strength in the rest of APAC and EMEA.

We therefore ended 2025 with a bigger backlog than in recent years. Note that none of these countries have the latest version of FSD Supervised available yet.\n\nOn the storage front, we had yet another record in terms of deployments.

I would like to thank our customers and Tesla in continuing this momentum. On the automotive margins front, automotive margins excluding credits improved sequentially from 15.4 to 17.9%.

The automotive gross profit was flat sequentially despite 16% lower deliveries, primarily due to regional mix, as we had proportionately more deliveries in APAC and EMEA.\n\nAs we look to 2026, with the progress that has been made with autonomy, our focus is on ramping production at all our factories. Our biggest constraint globally continues to be on the battery pack front.

While our teams have been creative in trying to resolve the situation by now putting 4680 cells in non-structural packs, we continue to iterate, improving things from here on.\n\nFSD adaption continued to improve in the quarter, reaching nearly 1.1 million paid customers globally. Of these, nearly 70% were upfront purchases.

It is important to note that beginning this quarter, we are transitioning fully to a subscription-based model for FSD. Therefore, net additions to this figure will primarily be via subscription model and, in the short term, will impact automotive margins.\n\nOn the energy front, we achieved yet another record in terms of gross profit for the quarter and ended the year with nearly $12.8 billion in revenue, a 26.6% year-over-year growth.

This was the result of higher deployments in all regions and continued strength in demand for both Megapack and Powerwall. As we look at 2026, our backlog remains strong, well diversified globally, and we expect increasing deployments with the launch of Megapack 3 and Megablock.

However, we expect margin compression from the increased low-cost competition, impacts to market from policy uncertainty, and the cost of tariffs.\n\nServices and other's margin declined from 10.5% to 8.8%, primarily from higher employee-related costs for service centers as we start preparing for the ramp in activity from the growth in the fleet size. We did see a momentum in margin We see We did see an improvement in margin from our Supercharging business, which is included within services and other.

Additionally, note that our robotaxi business-related costs, while not material, are also included within this. Given that this we're still in the early phase of our fleet deployment and are still doing a lot of validation testing, the revenue and cost per mile metrics are not meaningful to discuss at the moment.\n\nThen on total gross margin front, we ended the quarter with over 20.1%, something which we haven't achieved for over the last two years.

This improvement came despite the impact of lower fixed cost absorption and the impact of tariffs, which were in excess of $500 million in Q4.\n\nOperating expenses increased sequentially, primarily from increased stock-based compensation for employees and as we started recording charges on for one operational milestone under our 2025 CEO Performance Award that was deemed to be probable over the war term. Additionally, our spend on AI-related initiatives and new products like Cybercab, Semi, Optimus, and Megapack, etc., continues to be at elevated levels, and we expect this trend to continue for the full year 2026.\n\nNet income was negatively impacted from mark-to-market charges on a Bitcoin holding, which depreciated 23% as compared to the last quarter, and the impact of unfavorable impact of FX, primarily from our large intercompany borrowings.

On the free cash flow front, we ended up at $1.4 billion. We did end up CapEx being slightly below our previous guidance of $9 billion, but like as Elon has already mentioned, this year is going to be a huge investment year from a CapEx perspective.

At the moment, we're expecting that CapEx would be in excess of $20 billion.\n\nYou know, we'll be paying for six factories, namely the refinery, LFP factory, Cybercab, Semi, a new megafactory, the Optimus factory. On top of it, we will also be spending money for building our AI compute infrastructure, and we'll continue investing in our existing factories to build more capacity.

Then, also the related infrastructure along with it, and we'll also further expand our fleet of robotaxi and Optimus. While this may seem a lot, we believe this is the right strategy to position the company for the next era, and we'll make such investments, as Elon mentioned, in a very capital-efficient manner.\n\nNote that this does not include potential investments in solar cell manufacturing or a terafab, as we're still in early phase and we plan to provide an update in future quarters.

We're starting not the next chapter, but a new book on the progression of this company. 2026 year would be when all of this began.

While at times it feels daunting, it is going to be the most exciting change in Tesla's history, and we could not have even dreamed of embarking on this journey without the support of our customers and our investors. Thanks for again showing the confidence in us, and let's get ready for a future of amazing abundance.

Thanks.\n\nGreat.

Thank you very much, Vaibhav. Now we're going to head over to investor questions.

As always, we will start with questions from Say.com. The first question is, \"Today Today, there are approximately 90 million cars sold globally each year.

Does Tesla have a view, based on its robotaxi ambition, what this number will be in 5 or 10 years? How does this impact Tesla's EV strategy to have more models?\"\n\nYeah,

thanks, Travis. As Elon said, the future is autonomous, and obviously autonomy and Cybercab are going to change the global market size and mix quite significantly.

I think that's quite obvious. General transportation is going to be better served by autonomy as it will be safer and cheaper, and over 90% of vehicle miles traveled are with two or less passengers now, which is why we designed Cybercab that way.

In this new autonomous market, we at Tesla have the advantage of efficiency, cost, and manufacturing at scale that really no one else has. We've built that over the last decades.

We believe that that segment that we are creating will grow millions year-over-year.\n\nJust

to add to what Lars said there, the it's worth the the point that Lars made, which is that 90% of miles driven are with one or two passengers or one or one or two occupants essentially, is is very important one because that that implies that that the Cybercab, which is a dedicated, two-seater, dedicated robotaxi It's a little confusing with the terms robotaxi and Cybercab. Sorry about the confusion, but And in fact, in some states, we're not allowed to use the word cab or taxi, it's going to get even more strange.

It's going to be like cyber vehicle or something, cyber car. The the the the Cybercab, which is a specific vehicle model that we're making, does not have a steering wheel or pedals.

This is clearly, there's there's no, there's no fallback mechanism here. It's This car either drives itself or it does not drive.

We expect to start production in in April. As as always, it's a an S-curve of the production rate is an S-curve, starts off very slowly and then grows exponentially, then you hit the linear and then ultimately, it asymptotes at at whatever your target volume is.

We we would expect over time to make far more Cybercabs than all of our other vehicles combined.\n\nGiven Given that 90% of of distance driven or distance being distance traveled, exactly, no longer driving, is is is is one or two people, and I think it's like 80% is just one, it would mean that long-term, the Cybercab would would make several times more Cybercabs per year than all of our other vehicles combined.\n\nGreat.

Thank you so much. The next question, a bit related, \"Are there still plans to launch new models to address different price segments and vehicle types which could materially expand the TAM for Tesla?\"\n\nYeah,

to to further on what we were just talking about, we've launched our least expensive models ever over the last few months and are continuing to expand that the those models globally. Over the last decade, we have continually brought down the cost of our vehicles without sacrificing range, performance, or premiumness, and we'll continue to do that, as Vaibhav said, investing in our factories.

These are all trade-offs of where we spend our time or money. To Elon's point just now, with Cybercab coming, we are aiming to bring that Tesla premium ride experience to our largest market yet.

That could be 5 or 10 times our current levels of production. This new autonomous market, you have to start thinking about us as moving to providing transportation as a service more than the total addressable market for the purchase vehicles alone.

Of course, we do have plans to have robotaxis in various shapes and sizes, but obviously Cybercab will be the grand majority of that volume.\n\nYeah,

the the vast majority of of miles traveled will be autonomous in the future. I would say probably less less than I'm just guessing, but probably less than 5% of miles driven will be where somebody is actually driving the car themselves in the future, maybe as low as 1%.\n\nGreat.

The next question is, \"Historically, Tesla has spoken about gross margin per model. Are there standalone gross margin targets for the current models, excluding the benefits for FFD FSD sales?\"\n\nYou

know, we've talked about this with the previous two questions, but transportation, as we know, is changing, and I think we cannot keep applying the same framework from a car sales model to the future, what we are trying to do. It has to be looked at it more holistically.

In autonomy, software will be the driver for growth from now. As we aim to maximize the global fleet, we've been laser-focused on COGS from our side to make because that is something which we manage.

We will keep focusing on that, but I think we we need to look at it from a different dimension.\n\nYeah,

like the Cybercab the the whole design of Cybercab was it was to optimize the the fully considered cost per mile of autonomous driving. It's a different design problem than if you're trying to design cars for people who who will be driving versus being driven.

Cybercab is, like like said, super optimized for minimum cost per mile, and and also for a much higher duty cycle. We would expect Cybercab to be used probably 50 or 60 hours a week instead of the 10 or 11 hours a week that a a driven vehicle is used.

Typically, people will might drive their car for an hour and a half a day on average, it's like 10 hours per week out of 168. I I think an autonomous vehicle is likely to be used probably five times as often, which which means that you need to design the vehicle for a much more wear and tear per unit time and much more resilience.

It's more like a like a commercial truck that that's in continuous operation or close to continuous operation is is how you design a an autonomous vehicle.\n\nAnd and so so we will have larger vehicles than the Cybercab in the future that that are designed for full autonomy. We we've we've actually shown pictures of this, and in fact have shown prototypes, this is not exactly a secret.

In fact, we've given people rides in them, we're not we're not keeping this hiding this light under a bushel here, you know. It's like we're literally saying what we're going to do and have said what we're going to do for a while.

I think long-term we we would really the only vehicles that we'll make will be autonomous vehicles, with the exception of the next-generation Roadster, which we're hoping to debut in April, hopefully. It's It's going to be something out of this world.\n\nFantastic.

The next question we unfortunately have to skip because it's not related to to Tesla, and we would like to remind folks who use the Say platform to please focus these questions on on Tesla. With that in mind, we're going to move on to the next question, which is, \"What is the current bottleneck to increase robotaxi deployment and personal use unsupervised FSD?

The safety and performance of the most recent models Is it the safety and performance of the most recent models, or is it people to monitor the robot robotaxis in car or remotely, or is there some other blocker?\" I don't know, Ashok, if you wanted to kick off on this one.\n\nYeah,

we have scaled the robotaxi service that's available to customers over the last year in in order to just learn the scaling problems without having to like wait for the unsupervised. We see like two two goals.

One is like learn as much as possible from the fleet with the safety monitors, and secondly, be laser-focused with engineering team to solve the unsupervised FSD problem. I think we did both, like by the end of last year, we, we had a long tail of issues that we are able to churn through, and then in the last couple of weeks, we have started our unsupervised robotaxi service to public customers in Austin.

I think some customers took rides last week, and also the service continues today without any rear car or something like that. Separately, we did scale the fleet size in the Bay Area and in Austin, and through that, we learned, issues with charging and other issues that we would have seen once we sort of like scaled the unsupervised fleet.

Both are happening in parallel. A variant of the software that's used for the robotaxi service was shipped to customers with V14, and customers saw a huge jump in performance, like a lot of, happy feedback from customers.

Since then, we have improved the software significantly as well, and customers will continue to see with their own software releases that the software is so good that, they're like screaming to remove the driver monitoring software, because they're bored inside the car too much.\n\nAdding

to that a little bit with what Ashok said about learning about our charging and service needs, we're using our vast network of charging and service centers that really only Tesla has in this space to jump-start our infrastructure build-out needs to get ahead of robotaxi and autonomous vehicle demand. We expect that because of this network, we are the only company capable of scaling at the rate that is needed for the tsunami of autonomy that is coming.\n\nYeah.\n\nGreat.

Moving on to the next question. \"After the unveil of the Cybertruck, Elon stated that if it didn't sell well, Tesla would build a more conventional-looking pickup.

How practical would it be to create this new design on the Cybertruck architecture, and could it be conveniently built on the existing production lines?\"\n\nActually,

in its segment, Cybertruck continues to be a leader and is selling more than any other electric electric truck out there, while our competition continues to pull back. To the question itself, from a line standpoint, we always design our lines to be super flexible.

We built 3 and Y on the same line. We built S and X on the same line still, showing that we can do that.

The Cybertruck line was designed in the same way and is one of our most fully ready for

autonomy platforms.\n\nYeah. We will transition the Cybertruck line to just a fully autonomous line.

There's there's obviously a market there for cargo delivery, like say like localized cargo delivery within within a city, within a few hundred miles, something like that. There's there's a pretty there's a lot of cargo that needs to move locally within a city, and and an autonomous Cybertruck could be very useful for that.\n\nGreat.

Moving on to the next question regarding Optimus, \"Could you share the current number of units deployed in Tesla factories and actively performing production tasks? What specific roles or operations are they handling, and how has their integration impacted factory efficiency or output?\"\n\nYeah,

we're still very much at the early stages of of of Optimus. It's, it's an R&D still in the R&D phase.

We we have had Optimus do some basic tasks in the factory, but but as we iterate on new versions of Optimus, we we deprecate the old versions, and and so it's it's it's not I wouldn't say it's like it's not it's not it's not in in usage in our factories in any in a material way. It's more so that the robot can learn.

We we wouldn't expect to have, any kind of significant Optimus production volume until probably the end of this year.\n\nGreat.

Optimus Gen 3 is an awesome robot that

It is awesome.

It's an awesome robot that minimizes any differences It looks like a human. People could be easily confused that it's a human.

This helps our strategy for the AI too, because you can learn from how humans do the do these tasks, and it's very easy to teach the robot doing the same way, as opposed to previous robots.\n\nYeah,

I guess one thing I should say like is, there's a lot of news of like, various companies announcing layoffs and and whatnot, but, at at our Tesla in a factory in Fremont, we actually expect to increase headcount over time and and to significantly increase output from from our factories. We don't have any layoff plans.

We expect to actually increase headcount.\n\nGreat.

The next question, similar to the other autonomy questions, but slightly different, \"When is FSD going to be 100% unsupervised?\"\n\nWell,

it is 100% unsupervised in FSD is 100% unsupervised we we obviously have cars operating with no one in them and no safety monitor and no follower car or anything like that in Austin right now. For for customers, we we're we're being just very cautious with the rollout.

We with each successive version as we prove it out and we we make sure that there are no sort of unique issues in in in particular cities, because like sometimes you get like some very difficult intersection and it'll be an intersection where a lot of humans have accidents, by the way. It's like some some pretty nutty intersections where where there are a lot of humans make mistakes and have accidents in in various cities.

We want to make sure that FSD can handle those unusual intersections. Like you take LA, for example, where where Wilshire and Santa Monica combine is is like there's about I don't know, 20 traffic lights.

People are constantly having accidents there. You want to make sure that FSD can handle unique things in a particular city.

We're also just being paranoid about safety. With each successive release of of FSD, we will reduce the amount of driver monitoring that's needed proportionate to the safety of of the FSD build.\n\nGreat.

As it relates to robotaxi, what has surprised you about the rollout so far? We've talked about what's constrained the the fleet expansion to date, but it appears there are 200 vehicles based on public tracking.

Is that something that we can confirm?\n\nI

wouldn't say there is anything that like really surprised us because because we had a large fleet, we had all the metrics. There was not sort of a surprise.

It was just continued work to grind down on the long tail of issues, and that's what enabled us to launch the unsupervised service in Austin.\n\nYeah.

In in in terms of robotaxi vehicles carrying paid customers, I I I think we're well over 500 at this point between the Bay Area and Austin.

There's varying amount of like vehicles depending on the load, but yeah, you can have like more vehicles doing at peak times and then fewer vehicles in the off hours.\n\nYeah.

This will probably, double every month type of thing. It's It's on an exponential curve.\n\nI

mean, one other thing people forget that, we've been deliberate on all this in the sense that we have the supporting infrastructure already been in place, whether it's service centers, charging. Yes, we'll have to augment as the fleet grows, depending upon the density of where the demand is and whatnot, but it's not something like we just stumbled upon it and we are starting to.

We've been at it for years. Yes, not every city is designed the same way, same thing, our infrastructure is also not same in every city, but you have to give us credit that it's been a journey.

Like Lars said, if there's some company which can do it, we've already been at it, we should be able to deliver much better.\n\nGreat.

The next question is about chase cars, which we've already covered. Moving on to the last question, \"Elon, you've been spending significant personal time on Tesla's chip design.

Yeah.

What was the forcing function behind this increased involvement, and do you think external chip sales will represent a significant portion of Tesla's valuation by the end of the decade?\"\n\nWell,

I tend to spend time on wherever whatever the most critical issue is for the company, and and completing the AI5 chip design and having it be a great chip is arguably number the number one most critical thing to get done, which is why I'm spending more time on that than currently anything else at at Tesla. Spend pretty much every Saturday on on this, and and every and and and a chunk of every Tuesday.

It's like a if I'm spending my Saturdays on something, it's it's going to be something pretty important. I do think AI5 will be a a very good chip, and I I'm I feel quite confident about the design at this point.

Then AI6, which will follow that, it'll be aspirationally would follow that in under a year, will will be yet another big leap beyond AI5. I feel I feel pretty good about our chip strategy right now.

In terms of selling it outside of Tesla, we first need to make sure we have enough chips for all of our vehicle production and all of our Optimus production, and and then and then we will actually use the AI5 chips in our data centers. We already use the AI4 chips in our data centers, when we do training, it's a combination of the AI4 chips and Nvidia hardware primarily that we do training with.\n\nSo, but you say by the end of the decade, it's that's like things are changing so fast, it's hard to imagine like what happens at the end of the decade.

We we we might the when I when I look ahead at say what's what's the limiting factor for Tesla growth if you go say 3 or 4 years out, I I think it actually is chip production. Is there enough AI enough AI logic and enough AI and enough memory, enough RAM for for our volume.

Right now, I see that as being the the thing that probably limits our our growth in in 3 or 4 years, which would imply that we're we're not selling chips outside of Tesla, because we need them. In fact, I think I I think it's going to make sense and and this this is definitely going to be sort of a controversial thing, but I think Tesla needs to build a terafab.

I mentioned this at the shareholder meeting, but but even when we look at the output of the best-case output of all of our key suppliers, and I would say even they're beyond suppliers, they're like strategic partners like Samsung, TSMC, and Micron, the and and we say like what's the most you could possibly make, then it's it's not enough. We I think in order to remove the constraint, the the probable constraint in 3 or 4 years, we we're going to have to build a Tesla terafab, a very big fab that includes logic, memory, and packaging domestically.

That that's actually also going to be very important to ensure that we are protected against any geopolitical risks. I think people may be underwriting some of the geopolitical risks that are going to be a major factor in the in in a few years.\n\nSo, now, a lot of people like that's will say like, \"That's crazy.

Fabs are really hard.\" I'm like, \"Yes, I I know fabs are really hard. I don't think they're easy.

We do hard things.

we do a lot of hard things, you know. We didn't used to have car factories, and we didn't used to have battery cell factories, or lithium refineries, or Megapack factories, or, all these other things.

We we we figured it out.\n\nSo, I think it's I I think if we don't do the Tesla terafab, we're we're going to be limited by supplier output of of of chips. I think maybe memory is an even bigger limiter than than AI logic.

For example, we have we have chip supply deals with TSMC in Arizona and Samsung in Texas, but but currently there are no advanced memory fabs at scale in the United States. There are zero, literally zero.

Hopefully, Micron will have something going in a few years, because they are headquartered in Idaho, where they make a lot of potato chips, but we need to make computer chips, too. Anyway, we're working with our strategic partners on the chip front, memory and logic, but I think I think we we we've got to also try try our hand at building a a large-scale fab that integrates logic, memory, and packaging.

If we don't do that, we're just going to be fundamentally limited by supply chain, especially if there's some in a worst-case geopolitical situation, it would be quite quite a severe situation. I think we quite frankly, I think it'd be crazy not to try the terafab.\n\nSo, yeah.

Great. We'll have We'll have a

We'll have a bigger announcement on this in the future.\n\nAwesome.

With that, we're going to move on to analyst questions. The first analyst is Emmanuel from Wolfe Research.

Emmanuel, please feel free to unmute yourself.\n\nGreat.

Thank you so much. It's Emmanuel Rosner from Wolfe Research.

My first question is on the on the CapEx. You signal a pretty large increase to over $20 billion for this year.

I was hoping to better understand where the investments are going, any way to dimension for us which of the product line or technologies account for the bulk of the increase. Also, do you view this as like one-time in nature, 2026, or I guess how much of this is an ongoing level of high spending for for a number of years?

Then just finally still on that, with that level of spending, you're going to be burning cash. How should we think about minimum cash balance or any other way to to finance this?\n\nYeah.

Emmanuel, I tried to put this in my opening remarks too, but I'll try and go a little bit deeper. There's about six factories which we are starting production in this year.

There's a lot of cash, CapEx, which is going into that. Then as we are trying to scale Optimus, we need a lot more compute.

We're putting more money towards compute as well.

For training,

for training.

Yeah.

Then we're also going to be spending money to expand the capacity at, at existing factories. On top of it, we're just keep in mind that we're not none of these numbers which I shared of $20 billion factors in anything to do with the solar fab or the semiconductor chip fab.

Those would be, as Elon mentioned, would come later on.\n\nAnd you you think your second part of your question was, is this one-off or would we expect more? I think we're getting into this investment phase because we have big aspirations.

When you look at it, some of these aspirations are, I call them as infrastructure play. Especially if you have to do a chip fab and we have to do a solar cell manufacturing fab, those are infrastructure plays.

That that that funding takes a little bit longer. You you would be in an investment cycle for a little bit longer.

Initially, part third part of your question was how we going to fund it. Initially, obviously, we have over $44 billion of cash and investments on our books, we'll use our internal resources, but there are ways where we can fund it, especially when we look at the robotaxi fleet, because, any time any time you have a consistent stream of cash flow, you can go and get money from the banks.

We have had conversations with banks about it, and that is something how we're going to do it. Then on the infrastructure play side, yeah, like I said, we don't have a number yet, but given that it's it's an infrastructure play, it's a longer tail, we will have to look at a little bit more in terms of how we fund it, whether it's through more debt or other means.\n\nGreat.

Our next question comes from Andrew from Morgan Stanley. Andrew, please feel free to unmute yourself.\n\nGreat.

Thanks so much for for taking the question. I I just want to start on the the xAI investment that you guys announced today.

You talked about there being some collaboration, between the companies. I'm just hoping to get more more information or or if you're hoping you could shed more light on on what that looks like, and and maybe how the work xAI is doing can be leveraged at at Tesla and and vice versa.\n\nYeah,

if you if you look at the disclosure which we also put in there, we do talk about this is literally a furtherance of our Master Plan 4. Even today, if you look at Tesla vehicles, we are using Grok in there.

As we look at things whether we can do it ourselves, yes, there are a lot of things which we can do ourselves, but if there are things which xAI can help accelerate our progress, then why should we not do that? That is the reason why we've gone ahead with such an investment because this this is part of the strategic initiative, because as it is, if you remember, I talked about how many things which we're doing ourselves.

If there are ways and means we can find efficient ways for others to help us, and xAI literally fits into that mold, so that's why we went ahead with it.\n\nAnd

we just had like a lot of investors asked us to do this. There was like a lot of, investor Tesla shareholders said like we should invest in xAI, so that's like we're just doing what shareholders like asked us to do pretty much.

Grok will be, I think very helpful in say maximizing the efficiency of the management of the of a large autonomous fleet. If you've got a autonomous fleet that's, in the future 10 million vehicles or tens of millions of vehicles, then optimizing the efficient use of that fleet, Grok will be I think way way better than any heuristic solution or or sort of manually managed solution.

If you say managing say a large team of Optimus robots to build a factory, or build a refinery, and and say a rare hypothetically like a this is hypothetical example, a rare earth ore refinery, which we do desperately need in in in America, then you'd you'd say, \"Well, like what's going to organize the Optimus robots to to build that ore refinery?\" That would, you need kind of need an orchestra conductor. Then Grok would be kind of the orchestra conductor for the Optimus robots to build the hypothetically rare It might It might not be hypothetical in the future.

I'm just saying it's not not currently on our plans. We do we do need a lot more ore refining capacity in the US.

Then then what what's going to manage let's say let's say a thousand Optimus robots to to do\n\nUh

you're on mute right now, I'm not sure if you're trying to ask a follow-up question.\n\nAll righty. We're going to move on to the next question, which was is coming from Dan Levy at Barclays.\n\nDan,

please feel free to unmute.\n\nGreat. Great, thank you.

Elon, you talked about some of the constraints on memory. Given the very tight supply, are there any near-term constraints on procuring memory?

If there are, to what extent could you look at modifying the functionality in the vehicles similar to what you did in '21 when we saw shortages on MCUs, and maybe how are you thinking about bridging in the next few years?\n\nWell,

we the Tesla the Tesla AI is is very compute efficient and very memory efficient. I think one of the metrics one should consider for for any given AI model is the intelligence per gigabyte.

Especially when you're constrained on RAM, having a an AI that has very high intelligence density per gigabyte, you say like what for a given number of gigabytes, how much functionality can you get out of it? I actually think Tesla is ahead of the rest of the world in in intelligence density of AI by an order of magnitude or more.

Like it's this is going to sound like a pretty bold statement, but I'm I kind of know what the, what the intelligence efficiency of the of the big models are like Grok, and and like the honest and and and, bunch of the other models. Tesla's AI is like in terms of memory efficiency, more than an order of magnitude better.

That puts us in a pretty good position actually for for scaling. We we don't we don't We actually do think that there's we do have a solution for logic and memory for let's say the next roughly 3 years.

If you start going beyond 3 years, and we look at the scaling plans and how many fabs are getting built, and especially if you factor in geopolitical uncertainty, there's always there's always risk that maybe the those chips don't arrive that people were expecting to arrive. That's that's why I think we we need to have more fab capacity in the US, just in case, chips don't stop arriving for any reason.

This is this is really existential for for Tesla, because if Optimus is completely useless without an AI chip. It's not like at least the cars we can put steering wheels and pedals in, or retrofit them if if need be.

Optimus is just a mannequin without it's like the Tin Man of or from the Wizard of Oz, but but even worse. At least the Tin Man could walk.

Optimus wouldn't even be able to would just sit there without an AI chip. We've we've got a a good solution for a significant scale through, for the next roughly 3 years.

Beyond that, we we will be supplier limited, and so we've got to figure out some game plan to not not be supplier limited.\n\nGreat.

Our next question is going to come from George at Canaccord. George, please feel free to unmute yourself.\n\nHi,

everyone. Thank you for taking my my question.

There have been a surge of startups, particularly from China, entering the humanoid market. Wondering what the long-term competitive advantages that keep Tesla ahead are and and how, based on what you've seen, will Optimus fundamentally differ from these competitors?

Thank you.\n\nWell,

I I do think that the by by far the biggest competition for humanoid robots will be from China. China is incredibly good at scaling manufacturing, actually quite good at AI, as you can see from, the open source or not the open source, but the sort of I guess some of them are open actually.

Basically, the the the models that China is distributing for free are actually quite good, and they keep getting better. China is very good at AI, very good at manufacturing, and and will definitely be the the toughest competition for Tesla.

To the best of our knowledge, we we don't see any significant competitors outside of China, but China will definitely be be be tough competition. There's there's no two ways about it.

I always think like people are sort of outside of China kind of under- underestimate China. China is an ass-kicker next level.

I guess we're we're going to build, we think we think Optimus will be much more capable than any robot that we are aware of under development in China. We we think we'll be ahead in terms of the real-world intelligence, the electromechanical dexterity, especially the hand design, which is by far the hardest thing in the robot.

In fact, I'd say there's there's really three hard things about humanoid robots: the building a an incredible hand that that has the same degrees of freedom and dexterity as a human hand is an incredibly difficult engineering challenge, then there's the real-world AI, and and scaling production. Those are the three hardest problems by far for humanoid robots.

I think we're we're Tesla has is the only company with that actually has all three of those components. Yeah.\n\nGreat.

Our last question is going to come from Colin at Oppenheimer. Colin, please feel free to unmute yourself.\n\nThanks

so much. Guys, you talked a lot about the CapEx spend, but, this is incredibly ambitious technology development program that you're talking about.

Can you talk a little bit about the R&D spend and and how you're thinking about the synergies of of the different components, particularly on the hardware side, if you think about, batteries into into chips into into memory and and the efficiency of the system, and what sort of advantages you think you'll end up getting out of, some of these purpose-built devices that you'll end up integrating into multiple end markets?\n\nWell,

really all we're trying to do is make sure that we we can scale to very high volume with autonomous vehicles, with humanoid robots, and and that we address geopolitical risk, which I think there's so many companies out there that are asleep at the switch with regard to geopolitical geopolitical risk. They're like or or they just have their head in the sand and hope nothing bad will happen.

I'm I'm way more paranoid than that. I always think of Andy Grove's famous statement, \"Only the paranoid survive.\" And why did he come up with that statement at Intel?

Let's think. I'm I think there's a lot of wisdom in that statement.

We're going to be paranoid and make sure that we can continue to build batteries, and robots, and AI chips, no matter what happens. Companies that don't do that, a bunch of them will cease to exist.\n\nYeah,

remember, all this comes out of necessity. It's not that we want to do it.

It's just we have we have no choice.\n\nYeah.

We've built the most advanced lithium refinery in the world, by the way. It's not just like our lithium refinery in Corpus Christi, it's not just a copy of what others have done.

It's It's an entirely new process that is fundamentally more efficient and more advanced than anything else in the world. The same is true of our of our cathode refinery here in Austin.

We wish others would build this. Can can other people please, for the love of God,

Help.

In the name of all that is holy, can others please build this? Build this stuff.

It's not the first time we've asked.

Exactly. This is not the first time you've said something like this.

Like, why do we have to build these things? Why can others not also please Can someone else build these things?

It's a this is It is very hard to build these things. We build them out of desperation, not for any not because nobody else is building lithium refineries and cathode refineries.

We're pretty much the not just the largest, but also the only lithium refinery and cathode refinery in America. Yeah.

We're making the moves to make sure that, no matter what happens, Tesla will prosper.\n\nGreat.

Unfortunately, that's all the time we have for Q&A today. We really appreciate everyone's questions, and we look forward to talking to you next quarter.

Thank you very much, and goodbye.\n\nAll

right. Cool."

} Thank you. This concludes today's conference call.

You may now disconnect.