Tesla Q2 2026 Financial Results and Q&A Webcast

[00:00:00]Heat.

[00:00:09][music] Heat.

[00:00:19][music] >> [music] [music] >> Heat. Heat.

[00:00:38][music] >> [music] [music] >> Heat. Heat.

[00:01:05]Heat. Heat.

[00:01:17][music] [music] [music] >> [music] [music] >> Hey, hey, hey, hey, hey, hey, hey, hey.

[00:01:47][music] Heat. Heat.

[00:02:11][music] Heat. Heat.

[00:02:23][music] [music] Heat. [music] [music] [music] Heat.

[00:03:00][music] Heat. Heat. N.

[00:03:09][music] Heat.

[00:03:14][music] Heat.

[00:03:27][music] [music] >> [music] [music] [music] >> Heat up here. [music] Heat. Heat. N.

[00:04:05][music] Heat. Heat. [music] [music] >> [music] [music] >> Heat. Hey, heat. Hey, heat.

[00:04:54]Heat. Heat. N.

[00:04:59][music] Heat. Heat.

[00:05:13][music] Heat. Heat.

[00:05:21][music] >> [music] >> Heat. Heat.

[00:05:43][music] >> [music] >> Good afternoon everyone and welcome to Tesla's second quarter 2026 Q&A webcast.

[00:06:02]My name is Travis Axrod, head of investor relations and I'm joined today by Elon Musk, Viva Teneza and a number of other executives. Our Q2 results were announced at about 3 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.

[00:06:35]During the question and answer portion of today's call, please limit yourself to one question and one follow-up.

[00:06:41]Please use the raise hand button to join the question queue. Before we jump into Q&A, Elon has some opening remarks.

[00:06:47]Elon. [clears throat] >> Uh, thank you. So, yeah, it's been a great quarter. We achieved a record Q2 deliveries. Um, Model Y, I believe, is now I think it's the best selling car of any kind in the world and and is setting records across the board. So, it popularity is is increasing tremendously. And we're seeing in uh in locations that have um FSD approved, we're seeing a very high tick rate of of FSD. And and in fact, I think for a lot of people, they're actually buying Tesla full self-driving with a car attached as opposed to a car with FSD. They're they're coming into our stores uh in the US and saying they want they want the full self-driving and and uh with whatever car it comes with essentially.

[00:07:33]So uh clearly the this is this is a significant uh demand driver and um and as as we get approval for FSD and uh different part different countries I think we'll see a similar uptick in demand. Um so um yeah it's looking very good on that front. the energy at your business is also growing incredibly fast and um I think will be um crucial to for the scale up of artificial intelligence data centers. Um we're investing a lot in in growing the the core business and uh and and really preparing for the future. So this is this is a massive capex year. Um but but I'm confident that all the things that we're investing in are um what will yield uh incredible returns really maybe I think the best capex returns that that we've ever seen um and and just the you know Tesla team has performed incredibly well the sab has started production and we we also have uh started production and a ramping in uh well we we will soon stopped production with the with with Optimus.

[00:08:50]We have started production with the the Tesla semi-truck. We will soon stop production with Mega Pac 3. Uh we've started production with the lithium refinery, the cathode refinery. We also have we're scaling up battery cell productions, battery cell production, and we're also preparing to do a a massive uh solar panel like solar cells and and and panel production. Actually, this is going all the way from silicon refinement to producing the solar cell uh and then deployment of of solar because the there's going to be tremendous need for electricity in the future uh due to electrification of transport and uh and AI obviously. So, uh we're working on what we believe is the most ambitious buildout of advanced infrastructure manufacturing capacity uh ever uh in history. It's it's a big deal. So, with robo taxi. Um our goals are very ambitious for robo taxi but we do need to be cautious about um causing any accidents or causing any any harm to anyone. Um although there are I think 30 to 40,000 automotive deaths per year in the United States alone. Uh most of those do not generate any press when they you know there that they're you know you you never really read about almost any of those. Uh but if we injure even one person, it'll be worldwide head headline news. Um and regulators will immediately clamp down um on our our activities. So and we don't want to injure anyone. So So we're going as fast as humanly possible in scaling robo taxi um but with while trying to ensure that we do not harm uh anyone at all um and and ideally do not even run over a pet.

[00:10:39]So, uh, that that's really the constraint is we're we want to grow as fast as possible with Robo Taxi without harm to anyone.

[00:10:51]Um, and we're we're providing live updates for Robo Taxi launch as we open up in the city. Uh, we immediately post that on X. So, if you just follow the Tesla account on X, you you um, we'll keep you informed as we scale to new cities. Um, we've opened up in a number of cities in Florida and and in Texas uh and obviously in the in the the Bay Area. So, we'll we'll continue to scale uh I think very very rapidly on a on with um it looks like more than 10% a week in terms of of miles driven. So, it's a it's very high compound growth rate.

[00:11:35]So, let's see. Optimus. Um, you know, as you've heard me say before, I think Optimus will be the biggest product ever, but it is a very complex problem to solve. Um, it's it's one of the one of the hardest things to solve.

[00:11:49]Uh, to make a an an autonomous humanoid robot that can do tasks that you simp if you simply ask it to do something or show it a video, it can do the task without any programming.

[00:12:03]um like no one's no one's ever achieved this. Um and you've got there are many challenges in the electromechanical design of the robot uh to achieve sufficient dexterity and and reliably be very reliable and have a long long wear and tear. Like meaning it's it needs to be out in the field and not break down. Otherwise, you've got sort of a 70 kg robot that just flopped over and you you got to carry it like a carry it out like a body, you know, there's no wheels. So, and we don't want Optimus to go haywire. Um so, um it's it's a it's a lot of work to scale to get the the design right and to scale production. And and I really want to emphasize here that that the production scaling challenge is is is a very very substantial. This is going to be the hardest product to scale manufacturing that we we've ever made at Tesla because everything on the robot is new and the difficulty of scaling the production ramp is proportionate to the newness um of the of the parts in the robot. It's and you know at least with electric cars the non-electric powertrain portions of the of the cars had an existing supply chain. So you you could go to existing suppliers for wheels and side mirrors and windshield glass and and that kind of thing, body panels and but with Optimus there's there is no supply chain. So, we've had to build up a supply supply chain in its entirety um or in-house the production and we've we actually've in-house a tremendous amount. Um the Optimus production line that we we're building out in Fremont uh in place of what used to be the model SX production, it looks incredible. I mean, it's uh it's quite stunning to see.

[00:14:04]Um so um so I just want to be make sure to calibrate people correctly. It's octopus will follow the the sort of normal scurve of manufacturing of a manufacturing ramp but the initial portion of the x-curve will be will be quite flat and and long because of the newness of the um of the parts in the robot.

[00:14:28]Um, so, um, you you've probably seen lots of impressive demonstrations of robots on the internet, but but those demonstrations you're seeing are pre-programmed or remote controlled. So, there there is no humanoid robot that is actually able to do generalized tasks.

[00:14:49]Uh, Optimus will be the first one that is capable of doing that.

[00:14:54]um where it's not just a de not just a demo, it's genuinely useful in day-to-day life.

[00:15:00]So, and Optimus is designed to have um full human dexterity. So, a hand that has the same level of dexterity, if not higher, than a human hand. And the human hand is an incredible incredible thing.

[00:15:12]The more you study the human hand, the more amazing you realize how you realize how amazing hands are. There it's more more than just the posible thumbs. I mean the the the nuances of how human hands work are amazing as like the closer you look the more amazed you are. Um and um and Optimus will have that that capability. You'll have human and then superhuman dexterity. Uh so it's really going to be quite something.

[00:15:43]Um let's see. Terapab um you know the Terapab we expect to announce the location soon um and provide uh more details about our plans in that regard which but we'll leave that to the product sort of the the launch announcement rather than try to squeeze it into an earnings call because I think this is a this is a very big announcement and it deserves to have its own sort of uh day in the spotlight and and not be sort of squeezed into an earnings But I do think Terrafab is going to be an an amazing initiative and a necessary one and one without which we will will be constrained in our ability to scale Optimus production because we we simply won't have enough AI chips.

[00:16:32]So it's um it's it's crucial to solve that and we we'll we'll have to solve um memory logic and packaging for in order to scale Optimus.

[00:16:45]So um but anyway like I said we'll leave that to its own announcement. Uh we've placed equipment orders for our development fab in in Austin. Um and uh that development fab I think is pretty cool because it's intended to be to have um lithography mask production um and then logic, memory and packaging and chip testing all under one roof. So you you can have a very fast iterative cycle um and try out new chip designs very quickly and see if they work. I I don't think such a building exists anywhere on Earth. So, this is really going to be super helpful as we try some um exciting, adventurous, high-risk, high payoff bets on on AI chips. Um they focus, like I said, on primarily on Optimus, like how do we make Optimus as useful as possible?

[00:17:45]Um then we have like we have digital optimist which is basically um driving driving a computer screen. I guess you could call it computer use or something like that but driving a computer screen in the same way that you drive a car. So for the car it's it's sort of pixels in or photons in and controls out. The same thing is true for Optimus and digital Optimus. It's photons in controls out. So we we feel confident we can adapt the the same Tesla AI technology that we developed for uh self-driving cars to have a self-driving uh computer screen or self-driving computer essentially. Um, so the self-driving computer uh where where the the Tesla um the very low cost Tesla AI4 computer can uh handle all of the sort of short-term tasks. They're doing real-time video control of the screen of the computer. Uh so it's not like screenshots type of thing. It's real-time video um at high frame rate.

[00:18:53]And um and this get digital optimist will be important obviously for physical optimist because physical optimist needs to be able to operate computers like it can't come up to a touchcreen and not know what to do. It's got to have a generalized touchcreen and computer use capability.

[00:19:11]Um so that's that's looking promising.

[00:19:13]And um it this is in partnership with SpaceX. SpaceX is Brock sort of the big model that is kind of like the the manager of digital optimist and um and kind of tells digital optimist what to do. Uh it provides it with a series of tasks and then and then digital optimist goes and do does those tasks.

[00:19:37]Um so um we're also building out a mega pod design that has um Tesla AI4 computers with um with x86. This is a pairing um an X86 um computer with the Tesla AI4 computer in in a box. And it's kind of just digital Optimus in a box.

[00:20:07]um and and in in a mega pod, you know, kind of like the mega packaging.

[00:20:13]And so where we where we we put just a a large number of of AI4 plus x86 um combos in um in a giant box essentially. Um and these boxes can be placed anywhere in in the country or or outside the country. Um and this allows us to uh scale AI compute using um this aggregated uh electricity production. So you you because there's lots of places all around the world including at our superchargers. Um I think we've got something like seven gaws of power at our supercharges and and growing. Um and we can place uh megapods at at many of these supercharges and and have distributed uh power for AI. So I think I think that that could be quite a big deal. So, uh, in conclusion, we're super excited about our autonomy and robotics roadmap. Um, there's so much awesome stuff coming. It's, uh, it's, like I said, hard to squeeze into an earning school. Um, but we'll have a lot of product announcements. Um, and, uh, this is going to be a a great year for Tesla.

[00:21:26]I think one of our best years ever. And then I think next year will be even better. So, um, in conclusion, I want to thank the Tesla team for their excellent execution and all of our supporters for joining joining us in this journey.

[00:21:40]Thank you.

[00:21:42]>> Great. Uh, thanks so much, Elon. Uh, and now, Veov has some opening remarks.

[00:21:46]>> Thanks, Travis. So, Q2 continued the trend that we saw at the end of Q1, a resurgence in demand for vehicles across the globe. We achieved record Q2 deliveries globally with sequential growth. across the Americas, APAC and AMIA of 60%, 27% and 12% respectively.

[00:22:07]Additionally, Model Y set records in several key markets including the Netherlands, Australia, and New Zealand.

[00:22:15]The efforts of the Tesla team to create compelling products at a reasonable price with the cap capability of autonomous driving are coming to futition. We exited Q2 with our largest order backlog since 2023.

[00:22:31]We are therefore focused on increasing production at all our factories to meet this growing demand. Production growth will be limited by our supply chain.

[00:22:40]This includes not just batteries but also electronic components. We've been here before and like always, our Tesla team is working actively to unblock the obstacles ahead by securing strategic deals with suppliers.

[00:22:56]Like Elon mentioned, one of the key factors for grow for vehicle uh demand has been FSD. We our sales data suggests one of the main reasons comes are coming and looking at the car is because of FSD. In Q2 we had in North America about 55% of our deliveries had FSD subscription at the time of delivery enabled.

[00:23:23]Overall FSD attach rates continue to improve reaching nearly 1.5 million paid sub paid customers globally of which 55% is upfront purchases and the remaining 45% is subscriptions. We expect that the bulk of the growth in FSG monetization will come from subscriptions as we've removed the purchase option in most markets.

[00:23:47]Automotive automotive margins excluding regulatory credits declined sequentially from 19.2% to 16.3%.

[00:23:56]As a reminder, we had highlighted in Q1 that we had a 230 million benefit from warranty trowns and some tariff relief which did not repeat in Q2. Controlling for the impact of those benefits from the prior quarter, our automotive gross margins excluding credits would have been approximately flat. This demonstrates the effective pricing and cost management done by the Tesla team as we continue to drive demand.

[00:24:26]Commodity price increases and interest rate changes all continue to add to our cost. Note that interest rate subvention cost are recognized uh upfront as revenue offset. So as interest rates have risen this year the cost of subvention has risen along with them which had a negative impact on automotive margins.

[00:24:49]As we have previously noted, energy business is inherently lumpy.

[00:24:54]Deployments are tied to customer timelines and largely out of our control. In Q2, we deployed 13.5 gatt hour of energy storage, a 53% sequential increase and the second largest quarter for energy business. Despite this growth, energy gross margins declined from 39.5% to 20.4%.

[00:25:18]This margin decline was a result of several dynamics. First, there was a warranty true up in the quarter of about 240 million related to center vendor cell issues for our legacy deployments.

[00:25:30]Also, the tariff benefits we recognized in Q1 of more than 200 million did not repeat. And lastly, we had previously guided that ASPs for industrial storage is coming down amidst uh growing competition. Long-term, we believe the energy business should normalize at a gross margin rate in the mid to low 20% range. Our order backlog of this business is robust and we are doing our best to build based on both existing demand and future demand we expect from data center growth and overall electrification of the economy.

[00:26:08]Service and other margins improved sequentially from 9.2% 2% to 14.1% an all-time high. This was mainly driven by an uptick in volume and better cost management of our vehicle fleet that supports businesses such as used car, supercharging, service centers, and insurance. While marginal today, also included in this business are deliberate investments we are making in infrastructure that will help scale robotaxi in the future. We continue to grow the Tesla robot taxi fleet and have expanded to a total of seven markets in the US. We expect the ramp of the fleet to accelerate throughout the year along with expansion into new US markets.

[00:26:53]As previously guided, operating expenses increased sequentially. The increase came primarily from significant research and development related activities including pre-production ramp cost for new products like the semi-truck, optimus, cybercap and other AI initiatives as well as the appreciation for an additional compute that we brought online.

[00:27:17]Additionally, we had charges related to litigation expenses in the quarter. Note that we are in a big investment cycle and expect our operating expenses largely driven by R&D to continue to grow in 2026 and beyond.

[00:27:34]Net income was positively impacted by a marktomarket gain of 1 billion on our SpaceX holdings which was offset by losses on FX of approximately 300 million and on Bitcoin of uh about 100 million. As previously guided, our free cash flow ended up being negative for the quarter.

[00:27:56]Most of the reason for the it going negative is because capex more than doubled sequentially and we expect it to increase further in the second half of 2026.

[00:28:07]We continue to expect that capex for this year will be more than 25 billion.

[00:28:12]Capex will grow for the next two or three years as we expand our robot taxi fleet, expand our production capacity for Optimus, make investments for semiconductor fab, install solar manufacturing capacity and AI compute infrastructure. In addition to all the other expansions we'll do for other manufacturing for automotive, in addition to using our cash for such investments, we are being opportunistic in securing certain debt facilities that will give us the capacity to borrow up to 30 billion to help accelerate such investments. We believe this is the right strategy to position the company for the next era and we'll always make such investments in a very capital efficient manner. The path to amazing abundance is ever challenging and requires making bold bets. Our progress will not be will be nonlinear. The future is going to be great and we are ready to rise to the occasion. I would like to end by thanking the Tesla team, our customers, investors and suppliers for their confidence in us on this journey.

[00:29:21][clears throat] >> Thanks so much. Um, as everyone on the call may have noticed, uh, we've covered several of the say.com questions already in opening remarks, including Optimus production, Terraab, and Digital Optimus. Uh, Ashoke is joining us today, our VP of AI, and he's going to cover questions related to Robbo taxi. Ashoke, >> thank you, Travis. Many of the questions on say.com relate to robotaxi deployment and scaling. So, we thought we'd provide a comprehensive view of where we are now and where we are going in the future.

[00:29:52]First of all, I'd like to state that the robot taxi program has been operating extremely well, especially in terms of safety. The program has had an impeccable safety record. We have driven more than 380,000 mi of unsupervised robot taxi, now across six cities in two different states.

[00:30:12]We have had zero notable incidents.

[00:30:15]Any reports have been of other actors impacting us when we were stationary.

[00:30:20]Again, I'd like to emphasize how safe the operation has been so far. Zero notable incidents over 380,000 miles.

[00:30:28]Historically, the so-called experts have always claimed that you need lightars, radars, HD maps, and the entire kitchen sink to drive safely.

[00:30:36]Here, we show that such is not true. You can have safe, comfortable, and affordable autonomy with just cameras.

[00:30:44]This record should be a huge validation of Tesla's entire AI approach.

[00:30:51]Second, I'd like to discuss scaling.

[00:30:54]We started the robot taxi program roughly a year ago in Austin. We had safety monitors in the passenger seats of the cars back then. Around the end of last year, we had the first fully unsupervised robot axis in Austin.

[00:31:09]Since the beginning of this year, we have grown at doubledigit growth rates to the number of unsupervised miles that the fleet drives every week. I repeat, we have grown at such a high compounding rate on a week-overweek basis over the last several months.

[00:31:25]Not only that, we expect to continue growing at such a large rate through the rest of this year. The other thing to note is that for expanding to new cities, it has been relatively less effort on our front. We expect that the time to launch to a new city will continue to trend towards zero towards an end where we operate in entire states as a whole instead of going by city by city.

[00:31:51]Such growth was only possible because we have a world-class engineering team at the frontier AI development here at Tesla.

[00:31:58]The currently operating robotaxi fleet is already running early version early versions of the V15 FSD software that we had referred to in the past for V15. We had planned roughly about seven major improvement tracks uh and they're all happening in parallel. Uh the early V15 builds that are running on robot have already 40% of those tracks merged together. Uh and that's what's running in the fleet right now. As we continue to complete our work on V15, we will see that the car is going to be ridiculously safe and capable. We're already seeing internal signs that all of these bets that we have made for technology improvements are all continuing to are all starting to work.

[00:32:42]Uh and we expect that the payoff uh will be huge and that's what is going to help us sustain the same growth rate uh through the rest of this year and so on.

[00:32:55]Finally, regarding the Cyber Cap, first of all, it's a phenomenal product.

[00:32:59]Anyone who rides in it instantly falls in love with the experience. We have aligned our manufacturing targets to roughly match the projected growth rate of the unsupervised miles. The same V15 models that power the Model Y and other platforms will also work on cyber gaps.

[00:33:16]In conclusion, robo taxi growth so far has been literally exponential while keeping an impeccable safety record. Our technology bets are paying off helping us sustain the exponential growth for the foreseeable foreseeable future and the cyber cap is going to be awesome.

[00:33:33]>> That's great. Uh thanks so much Ashoke.

[00:33:35]Um and there was one remaining question on data collection for Optimus training.

[00:33:39]Would you mind covering that as well?

[00:33:40]>> Yep. Um regarding Optimus uh especially Gen 3, Optimus has been designed to not just match the appearance of human beings but also the functionality and dex dexterity of humans.

[00:33:54]One of the main reasons for this approach is that having the human form factor and function allows us to learn from humans on how to perform a wide variety of tasks. This opens up the entire world to provide data for training Optimus. Just like FSD, we have access to a broad fleet of humans giving us data from all of the workers at our factory. Optimus can learn quite a bit from observing them perform their tasks at the factories. We also have a dedicated data collection team who can provide relatively small amount but of very high quality demonstrations to do the post- training uh of these models.

[00:34:34]This is this is all on top of any training we can do off of the entire internet uh of data that shows demonstrations of people doing various tasks uh both in industrial settings and home settings.

[00:34:46]The second flywheel is when we have a large number of Optimus robots practicing their tasks in what we call the Optimus Academy. The data from the bots experiencing the task themselves will be invaluable and help us close any minor form factor gaps that may exist between the bot and the humans.

[00:35:05]This is also when the reinforcement learning loop kicks in where the bot initially attempts some task fails sometimes learns from both the successes and failures of those task and eventually learns to master those task at perhaps a superhuman level. Our AI strategy for Optimus is aligned the same. Uh it's the same end to-end strategy that drives FSD pixels in controls out. Um and just like FSD, we expect it to work broadly. In FSD, you know, you can get in the car, type in an address, and then hit start and it's like handles all of driving from park to park. And the same thing is going to be true for Optimus as well. You're going to just ask it to do anything. And then it just perform the entire task on its own without you having to do anything along the way. And it's the same team that produced FSD, you know, the V12, 13, 14, and now V15 software that now works on Optimus as well. Hence, I'm very confident that Optimus will be an extremely capable robot.

[00:36:00]>> Uh, great. Thank you so much, Ashoke.

[00:36:02]Um, so that covers uh the say.com questions. Uh, now we we will move over to analyst questions. The first question comes from Tom from RBC. Um, Tom, please feel free to unmute yourself when you're ready.

[00:36:22]Tom, you're still on mute. Uh, so please, uh, unmute yourself and go ahead and ask your question.

[00:36:35]All right, we're we're not getting any audio from Tom at the moment. So, uh we will try to move on uh to Andrew and we'll add Tom back to the queue uh to come in next. Um so, uh uh Andrew uh as soon as you can, please unmute yourself and ask your question.

[00:37:00]Hey, can you hear me?

[00:37:01]>> Yes, we can.

[00:37:02]>> Awesome. Uh, thanks so much for taking the question. Um, so maybe just to start out on Optimus, um, Elon, you mentioned that, you know, there's really no supply chain today for for Optimus, so you've had to mostly insource a lot of that yourself. I I'm just curious as as you've gone through this exercise, are are you seeing any suppliers willing to, you know, invest alongside you to to set up domestic manufacturing to potentially help you scale more quickly uh in the US. Uh you know, asking that in the context of obviously, you know, you want to control your own destiny to some extent and ramp as quickly as possible, but from an investment cycle standpoint, you know, potentially more capital efficient to have some partners, you know, help you along the way. So, I'm just curious how you're thinking about that and and some of the dialogue you've had with potentially some external suppliers.

[00:37:58]>> Elon, you may be on mute uh right now.

[00:38:00]We can't hear you.

[00:38:02]>> Sorry. Yeah. I our supplies have been have been great and they've they've made they have made and are making tremendous investments uh in in support of um Optimus and uh Roboaxi and whatnot. Um, you know, Samsung and TSMC in particular are um building FAS TSMC in uh Arizona and Samsung in Texas and putting in tens of billions of dollars uh for to build AI compute for um Optimus and and Roboaxi. Um Panasonic has also invested many billions in increasing uh battery cell production. Um, oh, I'm a little I'm a little I'm a little under the weather here. I'm a bit ill, so I I sound a little off. I'm kind of struggling with some I'm a little bit ill today. So, um, yeah, but but our partners have been been great. Um, and I'd like to thank them for their their support and their their investment and their hard work.

[00:39:06]>> Great. Thanks. Thanks for that. And and maybe just as a >> So, sorry, Andrew. We're going to have uh Kin hop in here uh for additional color uh briefly.

[00:39:15]>> Yeah, [clears throat] just to add to Elon's point that you know the the Samsung fab that's going to be uh pretty significantly dedicated to future projects. That's a massive investment multi-billion dollar. We're we're having we're seeing the same level of investment going into memory and also uh new specificities like metal injection molded parts, flexible printed circuits and all sorts of nonlinear technologies that are uh you know more based for the robot as opposed to the traditional vehicle supply chain. we've had uh and in certain cases where we don't find a great partner uh we we've never hesitated from endsourcing it and we have a very capable manufacturing engineering team and a design team that uh helps us scale those things.

[00:39:52]>> Yeah, I'd actually um I'd also like to thank Micron for um giving us memory allocation. um they they're going to make some very tough decisions on memory allocation and we really appreciate um Micron making room for Tesla um in the years to come and and giving us actually very significant allocation on on reasonable terms given the the pretty insane pricing of memory these days.

[00:40:19]>> Great. Really appreciate that that color and and maybe just my followup on robo taxis. you guys gave a lot of of great, you know, information in your prepared remarks, but you know, there's seems to be some evolving regulations at the state level potentially around sensor requirements. So, I'm just curious as you think about your own roll out, what do you want to see or what do you need to see from regulators to make sure that, you know, obviously some regulation might be good, but you don't want it to be overregulated. So, what are you expecting to see? What do you want to see to make sure that this is, you know, done in a proper way? and and we've got Lars on the line. Um Lars, if you wouldn't mind uh covering this one if you can.

[00:40:59]>> Yeah, sure. Um I mean I think you know we talk about regulation particularly in the US we've made a lot of headway um and you know just good initiatives from the the federal government in terms of the FMVSS rules uh moving towards you know acceptance and adoption of purely built AVs and I really appreciate the support from Nitsa on that. Um when it comes to like a state-to-state level, obviously the stuff in New Jersey um you know is a little bit disheartening. I think ultimately as a choke said in his opening remarks, we really focus on the performance of the vehicle and ultimately that's what's going to drive um regulators and the the public to to adopt it. You know, I think um you know, the best regulations are those that provide companies and innovators with uh a goal or a task and allow us to figure out the solution. And um when we see regulations that uh provide a solution before identifying the problem, we really don't, you know, appreciate those. But um as Tesla always has and will continue to do, we'll let our performance speak for itself. And um I think a show played that out really clearly and um that that that'll be the the foundation of our expansion moving forward.

[00:42:17]>> Great. Thanks for the questions.

[00:42:19]>> Uh thanks so much. Uh we're going to try Tom again uh from RBC. Uh Tom, can you unmute and try to speak, please?

[00:42:31]Unfortunately, Tom, we see you unmuted, but we're not hearing you. So, uh we're going to have to send you back into the queue. Um, next is Alex from Bank of America. Uh, Alex, as soon as you can please unmute yourself.

[00:42:46]>> Hey, uh, [clears throat] thanks thanks for the questions here. Um, just a few follow-ups on on robo taxi actually. I guess what milestone should investors be watching from here that would cause you to accelerate your fleet scaling or market deployments? you know, is it the further safety validation like a miles driven per incident or or something else? And then um would you ever consider thirdparty distribution partnerships such as with ride share providers uh to increase utilization or is the plan for now to keep uh robo taxi fully vertically integrated? Thanks.

[00:43:26]Yeah, I mean we expect to be vertically integrated with robo taxi as we are in the rest of our business. Um so I I don't think we're going to have any demand challenges with robo taxi. So this this is um the economics will be so compelling that I I think we will really have a lot more um desire to use the service than than you know like I think demand will strip our ability to service the demand. Um and and so obviously you just there's no no you can just go vertical in that situation.

[00:43:59]So it's really just about going through and it what we call the the march of nines of reliability where you need like like how many nines of reliability do you need to scale you know ideally you want 99.9999 you know percent reliable um and um and so that's that's the it's just the mines of reliability is I think the only thing really constraining our growth in roaxing Great. Uh, thanks Elon. Uh, the next question is going to come from Colin at Wells Fargo. Uh, Colin, please, uh, unmute yourself when you're ready.

[00:44:43]>> Oh, great. Uh, thanks for taking my questions. Um, there's been a lot of chatter about SpaceX and Tesla combining in some way. I mean, a lot of collaboration already today. I mean Elon, do you see synergies eventually from combining the companies? You does that make sense from you from your perspective over time?

[00:45:03]>> Well, I mean, as you can tell from all the many collaborations on on so many fronts with with SpaceX and, you know, and and there there's a lot there's there's more and more overlap. Um, especially with Terraab, that's that's really going to be, you know, a gigantic project. Um, So, but obviously, you know, we can't talk about, you know, combining [clears throat] companies and that kind of thing on on a school. It's it's got to be done with the appropriate process.

[00:45:32]And with that, I'll I'll turn it over to Brandon, our general counsel.

[00:45:35]>> Yeah. No, thanks, Elon. And that's exactly right. We continue to benefit from our relationship with SpaceX and we've had they've been a great partner and we have numerous beneficial transactions with them. And earlier this year, we deepened our relationship through an investment and a framework agreement. This will allow us to continue to work with them on projects that Elon mentioned like Terraab and Digital Optics.

[00:45:58][clears throat] >> Yeah. And then there's there's many other things too. Obviously, you've got you've got Grock in the car. Um and um so and and Grock helping yeah drive digital Optimus. Um, you've also got Starlink being integrated into um the uh Cyber Cap and and Sonic will be integrated into all of our vehicles at least for markets that uh Starink is active um because for a rover taxi situation um you need to have you know coverage everywhere and there are many many places even in Silicon Valley where the cellular coverage is is a ter terrible or sometimes non-existent which is surprising for Silicon Valley but um but I know when I when I drive to work those first 10 15 minutes I I can't actually do any calls because the cell connectivity is so bad. So we can't have uh robo taxis um getting stuck in these like Bermuda triangles of lack of cellular connectivity. Um so so Starlink with its ability to do connectivity anywhere is actually quite important.

[00:47:04]Um, so we don't have, you know, car robo taxis missing an action. Um, and then obviously if if if people are sitting in the car, they're they're going to want to do high productivity stuff or or entertainment. And with Starlink, you can watch, you know, full 4K live sports um in the car and with with very low cost per gigabyte of data as that's that's really not feasible via the cellular system. And there's many I think there's many other situations.

[00:47:36]>> Yeah. I mean I would say think of it like uh Ashok was mentioning about cyber cab right that the experience is great in a world when you don't have to focus on driving you have you have all the play all the time in the car to do whether it's a conference call watch a movie or anything. That's why if you look at the cybercap design, it has a big screen and you know we've [clears throat] started doing cybercap rights in uh our factory in Austin and when you go through that experience you I mean soon customers would be able to do it in not too distant future and then you'll understand why that connectivity becomes so important.

[00:48:18]>> Great. And then Colin do you have a follow-up? Yeah, just uh if I follow up on on robo taxi, you know, if I look at the launches, it's been you've been adding cities, but the the number of units is still, you know, looks like based on media reports sort of in the dozens as opposed to hundreds. You know, why not just sort of scale up Austin or one or two cities before adding cities and and sort of what do you need to sort of get that, you know, higher volume numbers in in a major city? what what sort of is the roadblock to sort of adding more vehicles on the ground?

[00:48:55]>> Yeah, the reason we have been expanding across different cities instead of just doubling down on a single city is that we want to make sure that our stack is a very general one. We it is a general one. We just want to like you know both pro to ourselves and to other folks that it is uh working across a lot of different cities without too much effort per city. Uh and that's what we see internally. uh and then as Iron mentioned earlier the growth rate has been uh literally exponential just it's in the early part of the exponential that's why it's hard for others to comprehend in terms of like miles versus the vehicles since these uh vehicles operating the robot taxi fleet drive basically uh continuously as opposed to human drivers who who use vehicles for maybe a couple hours a day or something like that. These vehicles are in mostly continuous operation which means that even for a few vehicles you can get a lot of miles out of them and that's why we are tracking the amount of unsupervised miles or just the number of vehicles or something like that and we want to be operating in a very efficient fleet. So all of the work that goes into uh the efficiency makes it even more so that we get a lot of miles per vehicle available in the fleet.

[00:49:58]>> Yeah. And just to add further to this, I'm like what think of it, we're trying to scale this and we're trying to make sure that there are different kinks which would be there and there are things not just on the software front on the operation front which we're also trying to tackle. So that's why we're trying to have a broader footprint to try and make sure before we go really high in terms of deployment that we are able to sort these things out in a smaller fleet in a controlled manner.

[00:50:31]>> Yeah. One of the things for cyber cab should be noted is because it is a new vehicle chassis um we need we need to accumulate um driving data that is specific to the cyber cab before we can put a lot of them on the road. Uh so unlike say Model 3, Model Y and our other vehicles where we've got a lot of lot of vehicles on the road, millions of vehicles on the road, um we we we don't have that for subcap. So we actually have to accumulate um miles with sold caps that are retrofitted with um steering wheels and and brakes you acceleration braking paddles that kind of thing um to calibrate to the the cyber cap chassis. So, so you'll see uh as as as we are confident about that the number of cyber caps and cities will increase dramatically.

[00:51:28]>> Great. Our next question Oh yeah, go ahead.

[00:51:31]>> I can just add like one thought here.

[00:51:33]The the other reason I just want to like point out is you know transport like TCO transportation regulatory uh situations are different city by city state by state is not a federal you know framework to do that and so like the reason we're expanding city by city is to make sure that we're meeting all of those one at a time and we're satisfying you know all of the requests of each individual city and and you know in in every way we can. So we have to expand, you know, to each city to meet those requirements and we'll continue to do so and and and to all of what Ashok and Vib and Ashok said. We'll we'll follow up the skilling.

[00:52:11]>> Thanks Lars. The next question is come from Walt from Lightshed. Uh Walt, please unmute yourself.

[00:52:18]>> Thanks. Just a quick followup for that Lars comment. I mean the Nits administrator Jonathan Marson was just on CNBC. you know, he already, you know, was trying to get rid of the pedal and said, I forget his exact term, but seemed like he wanted to get rid of the steering wheel. It seemed like that would be a gating factor, but is there other things that you need at the federal level to really kind of unlock um your ability to to ramp cyber capab?

[00:52:46]>> Um I I mean the short answer is no. I mean, I think we have a great relationship with Nitson and and you know, director administrator Morrison especially. Uh, and you know, I think they're just following through with with what, um, you know, the public wants and the world knows is coming and they're trying to stay ahead of it and and make sure that, you know, they're they're doing the right things in place. So, um, you know, we've been open and honest with with them for the last couple years about what our plans were and what we're doing. And um you know we don't want to say we're in lock step but but uh you know I feel like we have a partner there and we're working together on it.

[00:53:22]>> Got it. And then just Elon kind of a roadmap question I guess. Um Starlink obviously is integrated in Cyber Cab and and you guys just talked about the connectivity for watching movies but what about the flip side of that? Can these can the Cyber Cabs become remote hotspots for Starlink mobile? And then similarly roadmap question on the semi you know when do you anticipate um looking at autonomy for the semi because that seems like a pretty sizable market as well in terms of autonomous trucking and the need for it given what's going on in our country right now.

[00:53:56]>> Yeah absolutely there as you um allude to there is a a really serious shortage of truckers. Um so um there just aren't enough people who around who who want to drive trucks um which are crucial for transport throughout America. Um so so an autonomous semi is actually going to be very important to address the the shortage of truck drivers. Um and and obviously will be great for um improving safety and uh and you know making it easier on truck drivers to have a self-driving semi is is going to yeah I think dramatically improve safety and and comfort for truck drivers that um you know yeah use the Tesla semi. Um the you know since the the units the total number of units of the Tesla semi is still low and and will continue to be a very small percentage even by the end of this year of our total vehicle fleet.

[00:54:56]U it makes sense for us to focus our efforts our self-driving efforts on our high volume uh vehicles. So, Model 3 and Y um and and solving self-driving for and really getting, you know, the point where it's a generalized unsupervised self-driving for uh for for those vehicles um and Cyber Cap. So, so we expect to um do get get self-driving working on the the Tesla Semi um probably around the end of this year or early next year.

[00:55:31]Um and um but we I just don't want it to be uh a distraction um on the you know on the march of nines of of um self-driving of of safety for self-driving on three wine cyber cap that's uh so that it's uh taking a bit of a backseat for the next 6 months or so for autonomous semi but it will be um you know it will definitely be a bit working next year and um in in time for the scale up to high production of the Tesla semary um for for for communications. Yeah, I think I think there's an interesting angle there where um you you could use the uh Starling terminals that are in the Tesla cars as um [clears throat] like sort of cell towers I suppose or connectivity towers basically for providing relays on the ground to um to cell phones and to um anyone who who would like Wi-Fi basically. Um but yeah, I mean this is a different subject but obviously we could do that with the stationary stalling terminals too.

[00:56:50]>> All right. Um our next question is going to come from Will at Truist. Uh Will, please unmute yourself.

[00:56:58]>> Great. Thanks so much for taking my question. Um, uh, Elon, you teased us a little bit about, um, sort of the supply chain relative to Optimus, highlighting how difficult it's going to be, but we'd be really interested to understand what your propensity is uh, to um, manufacture uh, semis for this yourself. When we think about microprocessors, microcontrollers, actuators, and other uh semiconductors that are going to be used, it it occurs to me that uh you have many that are commercially available. You have an option to design some but have them uh made at at third party fabs. You also have the option potentially to design and make them yourself. What what should we think about your propensity to uh source these components in these different manners?

[00:57:49]you're talking about like um making like microcontrollers and small like the various computers that are in the car that operate um you know like the steering system and the braking system that >> I'm I'm talking about for Optimus not for um >> yeah um Optimus has a lot of um really specialized power electronics and and circuit boards um it's all Tesla design um design stuff. Um but but it is fabricated by suppliers um for Optimus uh 4 um which will be built in um Austin um that that will be a much more vertically integrated uh supply system for Rockimus 4. Um and that and that would you know we'd aim to have an order of magnitude more production of Optimus 4 than Optimus 3. So sort of aspirationally 10 million units a year versus a million units a year of Optimus 3. Um [clears throat] um but but you know with all the caveats there which is insanely difficult to scale production.

[00:59:00]Um so um and and yeah Optimus Optimus 4 will be much more vertically integrated.

[00:59:06]So probably doing a lot of the sort of PCV work in inhouse for that.

[00:59:13]uh as a followup um well first maybe you could talk about the timing as to when that when Optimus 4 should be um should let's say start production but the the actual followup I wanted to ask was uh the plan to upgrade hardware 3 to hardware 4 in order to uh be able to uh process um uh I suppose it's version 15 of the and and future versions of FSD.

[00:59:40]Is that still the plan to do hardware upgrades on the cars?

[00:59:45]>> I think it's going to make sense for for to upgrade um all cars with that are um have less than hardware for almost all any cars that have cameras basically cuz otherwise it probably too too many modifications. But anything that's got sort of um that that's that's set up for cameras uh it'll be financially sensible at some point to upgrade them. Uh and but I think we'd want to upgrade them to the next generation of um you know like next generation AI AI board. So it would be either um like we haven't we have an upgraded AI4 uh chip that's sort of a medium sort of a moderate improvement over AI4 probably reaches production around the middle of next year. Um and then there's also AI5 which hopefully is in volume production um around the build of next year. Um AI5 will initially go into Optimus.

[01:00:45]Um and uh actually making very good progress. So the Tesla chip team is really doing doing great work and making incredibly fast progress on AI 5 and and we're I'm very excited about the design of the Tesla AI6 chip. I think it's I think it's going to be the best edge computing chip in the world. Um, and if somebody's got a better one, I'd like to meet that person to shake their hand cuz the AI sex chip is just so good. Um, so I think we got we got a I think it's going real well, really well on the chip front. Um, yeah, and again, I'd like to thank our thank TSMC and Samsung and Micron for the support.

[01:01:29]>> Great. Uh we are coming up on the hour, but we're gonna try to squeeze uh Dan in. So uh from from Barclays, uh Dan, please unmute yourself.

[01:01:42]>> Uh great. Thank you. Uh appreciate it.

[01:01:45]Um I'd like to ask about uh the the pace and path of capex. Elon, in the past you've said that, you know, you'll spend as much as you can, but you know, really efficiency seems like it's the gating uh factor. So to what extent right now in this capex cycle is your pace of capex effectively capped simply because spending any more would be done in an inefficient manner? And to what extent is the the pace of capex really the primary determinant right now in unlocking the different supply constraints that you have?

[01:02:21]>> Yeah, it's a good question. Um I I've actually you know my my what I've asked the Tesla team to do is is like really we should be spending on capex as fast as we can spend as fast as we can without without it being too wasteful.

[01:02:38]So, I'm not trying to aim for like some extremely high efficient efficiency capital spend because that would slow things down. Um, so it's trying to it's a balance between like how much cap, you know, sort of capital efficiency versus time. Um, so if if we can be it's okay to be a little less capital efficient if we get things done sooner because that's actually going to be the the higher um MPV outcome for the company. Um, so, um, overall I'm pretty happy with how things going. I mean, we're doing an incredible amount of construction and and production growth in so many different arenas simultaneously.

[01:03:21]Um, I don't think there's ever been a company that's done this to at this scale. You know, maybe proportionately back in when Henry Ford did the Model T or something like that, but then or World War II when uh, you know, companies were were making tattoos or scale up weapons production but I think probably this is the fastest industrial scale up and since World War II in America.

[01:03:49]>> Yeah. And I will just add a little bit on that. Basically, if you look at it, you know, all our spend on capex is trying to do productive assets, right?

[01:04:01]We're basically scaling factories, whether it's Optimus, whether it's CyberCap, whether it's LFP factory, which went into operation earlier this year, or the semiaactory or even putting in a semiconductor fab in there, as well as, you know, we are trying to get a lot of solar manufacturing in the US. In fact, >> yeah, yeah, this is underrated, by the way. Please go ahead. Yeah, I was going to Yeah, basically you know the amount of solum manu infection which happens in the US we're just going to uh multiply it with an order of magnitude and none of these things are easy right because what you have to do is in order to build a factory you have to start from ground zero so just so you know again I don't think we have shared this we end up being the general contractor for almost all our construction facilities because we are doing that much construction and >> so going back to what Elon was saying we're going at a very rapid scale and yes uh you know this means that we have we are doing a lot of things all at the same time and that's why you know we just have to go as fast as we can humanly make uh things work in the real world >> yeah our capex efficiency I think is offscale Great.

[01:05:29]>> All righty. And then Dan, yeah. Did you have a follow-up that we can wrap up with?

[01:05:33]>> Yeah, maybe just a quick question on on energy storage, which uh you know, I I think is uh uh a key topic here. Maybe you can just address you know, to what extent is energy storage going to be supply constraint for the foreseeable future? And then just to what extent are are the deployments right now the typical peak shaving that you'd see at utilities versus you've talked about in some of your materials um you know for battery storage addressing the the quality of power issue for data centers given the sharp fluctuations in power use where you have an advantage given software power electronics. So just what is that demand curve right now between data centers versus just broader peak shaving for utilities?

[01:06:16]Um well it's still mostly um it's not just peak shaving but it's also grid balancing. So it's not it's not just like the facts are on briefly but for balancing for balancing power from um wind generation and solar generation uh batteries are are awesome. Um I mean the solar battery combination will be how the vast majority of energy in the world is produced in the future. Um, that's already of course how all of the Starling satellites are are powered is with solar solar panels and batteries.

[01:06:50]So, and you can just think of Earth like a giant satellite.

[01:06:54]Um, there's there's far more power available from the sun than anything else by far. Um, and uh and we we think power constraints are going to be there already a major issue for for AI, you know, just turning on the the AI computers. AI comput demand is so high that um people are having even the so-al you know the hyperscalers are having trouble turning on the their their um their AI compute and finding the power and then um smoothing the power with uh you know especially for the training runs where the power cycles dramatically um in a in a in a very short period of time you can have during a training run the the power consumption can drop by you know 70% for 100 milliseconds.

[01:07:39]Um, and you really need fast acting advanced power electronics to be able to um smooth out these these um massive the massive massive changes in power um especially during during their training runs. So um that's why SpaceX has bought so many mega packs for the data centers. It's actually um um mostly for for smoothing out the power for the training runs. Um but will also be useful for um being able to to have more power from the um the grid by uh you know if you if you tell utilities that they can um you know that that that they will they will not need to supply power during the you know the worst hours of the worst days of the year because the batteries can handle it. then that makes it much easier for that for them to um to to give you power. Um in fact, I mean the probably the best way to scale total energy production in the United States is simply using batteries because there's I think around 1.2 or 1.3 terowatts of power production in the US, but the average power usage is only half a terowatt. So you've got two and a half times as much power generation as the average usage.

[01:09:04]um that means you could potentially double the energy up for the United States just with batteries.

[01:09:11]So really we see demand for mega pack being very very high in the future.

[01:09:21]>> All righty. Uh that is all the time we have for today for Q&A. Uh we appreciate everyone's questions and we look forward to talking to you next quarter. Thank you very much.

[01:09:29]>> Thanks everyone.

[01:09:33]>> [music] [music] >> Heat.

[01:09:44]Heat.

[01:09:48][music]