Rivian Automotive Aktie

Hello, everyone. I'm John Saager, Director Global Autos here at Evercore ISI. With me today is James Philbin, the SVP of Autonomy and AI for Rivian; and Chip Newcom, the VP of IR. James and Chip, I'll leave it to you guys to kick it off. Maybe give us an overview of Rivian and some of the key talking points you've been dealing with the investors.

Yes. Well, first, thank you so much for hosting us today, and great to be here to be able to talk to investors. So -- so in terms of the key areas, I can talk a little bit and then James will pass it over to you on autonomy. But really what our investors are focused in on right now. First is -- sorry, I'm getting some feedback all about the ramp of the R2 right now. So that's our midsized SUV that we launched back in June. And so as you can imagine, this Q3 being the first quarter that we'll have really had a full quarter of production, lots of questions that I have no doubt will talk about how is the ramp going from a production and a supply chain perspective, the targets that we laid out on the Q2 earnings call, and then, of course, a lot of focus on autonomy and our path there.

Yes. So yes, the team has been laser-focused on getting point-to-point ready this year. So we're very much on track there for initial rollout towards the end of this year. And we've really been on this sort of autonomy journey since the launch of Gen 2. And Gen 2 was built around beside our vertical integration in a very strong data loop so that we can collect customer data, customer sensor data together with the driving action. And that's really what's allowed us to drive the autonomy developments. We launched Universal Hands-Free last year, end of last year. That's the nation's, like, largest-scale hands-free system that you can get. And with point to point, we really think this will be the next step, important step in our autonomy journey as we iterate towards those kind of higher levels of autonomy in the stack.

Maybe you could talk about the development of your AV stack over time and to date and why you think multimodality is the correct approach?

Yes. So we believe very strongly in multimodality. So we think that the -- even if you accept the premise that cameras could be comparable to a human level perception. We think there's many cases where actually that's not enough where humans also make mistakes. And we think that the sort of societal bar will almost certainly be better than a human level for widespread autonomy. And so yes, for those reasons, we think that, yes, multi-modality really helps in the field.

It also helps us accelerate our progress. So what we see is that when we have radars with cameras, it actually makes the system that much better. When we add LiDAR, it makes it better again. And so this is really a way to accelerate our progress as we're building the scale with R2 as well as providing a system -- a set of sensors that really provides a safe and performant and robust capability to our customers.

And how has that developed over time?

Yes. So we started, let's say, 3 years ago, where we brought the full autonomy stack in-house. So we -- everything from the lowest level camera drivers and ISPs all the way to the perception models running on the vehicle. So Gen 1 had a Mobileye system. With Gen 2 when we launched that, we pulled that system out and replaced it. And then as we've been gathering that data developing our perception models, we've been steadily improving first of all, with a Highway Assist feature then Universal Hands-Free and now the big focus is on point-to-point.

What are some of the technical hurdles that are still remaining for you guys as you look to roll out point-to-point by the end of the year?

Yes. So autonomy is really about kind of the edge cases. And so a big focus for us has been not how can we get more boring highway miles, right, which are not very informative for the model, the sort of gradient you get back when you train with that data is very, very minimal. But how we can find more interesting cases, which are really sort of driving kind of different behaviors from the models, so the gradient is large. So as part of that, we developed this grid system where we can actually using VLMs, modern AI techniques, we can scale up the collection of edge cases and so we currently have like a 3,000 query set of very long tail edge cases that are constantly querying against the customer fleet.

And then we're using that to train the model. So I think that's a big part of the development process is actually getting those edge cases getting new things that the model hasn't seen before so that it can refine its driving behavior.

So you had mentioned point to point by the end of the year. Where does the autonomy program go from there? What are the next steps and ODDs that you're looking to unlock?

Yes. So we'll begin the rollout of point-to-point at the end of the year to probably our early adopter customers, and then we'll continue that rollout into next year with the aim of having it rolled out to the full Autonomy Plus customer fleet later next year. As obviously, we're rolling that out. We'll continue developing and improving the system. So we'll get improved disengagement rates, more scenarios with better handling. There's also some features. We probably will have out of ODD for the first version that we'll aim to add in, for example, parking lots.

And then a big focus in on the Uber robotaxi partnership, where we're then iterating on that system getting it to a fully driverless level of performance. And that's both on the model side but also on the validation side, so increasingly leaning into simulation, large-scale simulation to measure and improve the system.

And the Uber robotaxi deployment, will that be domain defined in certain specific roads and areas? Or will that be something that you can use sort of everywhere?

So yes, the system is really like a single stack. So we will be training the model to handle everywhere. Now it may have fine-tuning data that is specific to particular cities or ODDs. And we've actually started building that through expert fleet deployments in our first 3 cities, San Francisco, Chicago and Miami. And we actually have an expert fleet driving around here right now.

So all of that data will be fed into the model to really help kind of get that expert that Rivian driver policy dialed in, especially for point-to-point urban driving dense city driving.

And will the first point-to-point release be Gen 2 hardware only? Or does it require Gen 3?

No, it will be the whole on R1 Gen 2 and our Gen 2 Phase II and Phase III fleets.

Okay. But where drivers, Hands-Free, eyes-off where drivers don't really need to be actively involved is sort of a major milestone in the industry. What's the largest gating factor for you guys to get there? Is it validated miles, regulatory liability sign off? Or is there some sort of technical hurdle that we're having?

We have Yes, there's probably 3 pieces to it. I think one is just raw model performance as we're improving the model performance of say in the long tail, that will help. There's then the validation piece like how you can actually measure that model performance. And you really don't want to do that by having to drive every release, you want to do it offline in the cloud. And then the third piece, yes, is regulatory. Currently, it's a complete patch work in the U.S. So a lot of states view eyes off essentially in the same bucket as L4. And so that means in some states, you can ship that system, others, there is a pathway, but you have to jump through a certain number of -- go through a certain regulatory process. And in other states like New York, currently, there is no pathway. So it's a patchwork.

So you've got sort of navigate that.

Yes.

Playing hopscotch.

Yes, definitely.

So paid autonomy is the $2,500 upfront or $50 a month launched in April and you guys have talked about that being encouraging take rates to date. Can you frame the attach rate on new deliveries versus the installed base? Is there a gap or a difference there? And then the sort of upfront versus the subscription mix?

So we haven't quantified that as of yet. For everyone who's buying an R2 right now as part of the Launch Edition package, they are getting Autonomy Plus as part of their purchase. So we'll see as we shift beyond the launch edition package, what that take rate looks like for the R2. But as you pointed out, for the R1, it's been better than we expected.

Yes. Can you just frame out how much Autonomy Plus contributes to your software revenues?

It's still relatively small, just given the size of the car parc and the monthly price or the sort of full price if someone is purchasing it upfront.

Okay. And then RAP1 development is sort of on track and in the final testing phases of the production. So like, what's the BOM delta of the dual RAP1 plus all the sensor suite, the 11 cameras and radars? Is that cost carried in the R2 margin? Or is that offset by autonomy costs sort of pricing?

So from a revenue recognition perspective, it's going to be recognized within autos and auto COGS. So there will be a cost around putting LiDAR into the vehicle. Now the cost of LiDARs has come down significantly. So you're talking on the order of a few hundred dollars versus if you went back several years ago where it was thousands or even tens of thousands of dollars within the vehicle. So that will be beared within the cost of the vehicle. And similarly, as we launch our Gen 3 computer in the vehicle, we actually see a benefit in the sense that we're no longer paying away that margin to our silicon provider. It will be our own in-house silicon.

From an autonomy perspective, how do you think about your position relative to the competition, most specifically Tesla, but also Waymo and all the other players that have spoken at the conference today. Where do you think you fit? And how do you think you either catch up or surpass some of those other players?

Yes. I think we're -- I mean, a, we're vertically integrated. So I think we're both building the vehicle and the autonomous stack. And I think we have the benefit that we have a very large shipped car parc already collecting data. So that means that we get huge diversity already from our customer fleet vehicles. I think in comparison, a Level 4 company like Waymo, they actually have to go city by city to really get that data and do their validation we get the sort of benefit of being able to train the model and from the get go with all of that data.

I think the second is that we are really leaned into like the end-to-end approach. We think that is the right way to build the autonomy systems longer term. amazing when you see the capabilities that kind of emerge from that system. I'll give you an example. So narrow negotiations, you're on a narrow roads, 2 cars coming and someone needs to pull out. So I actually worked on these systems at Level 4 companies in the past, very, very complicated set of heuristics or the things that you have to put in place to get that to work.

And this model kind of -- it just falls out of the model. So there's nothing extra needed -- so I think when you see things like that, you realize, okay, this really is the right way to build this system the right scalable approach. And then -- but in comparison to say, Tesla, we really believe in multimodalities. And don't think there's really a tension between training end-to-end or and using multimodal sensing. The model sees every sensor essentially the same way. It's a bag of numbers and it encodes them using its own encoders. So it's able to figure out those cross correlations and how to best benefit the output using its inputs.

Right. You mentioned the large and growing fleet bringing in data today, how do you filter through what data is useful and what data is not? I mean, you mentioned the 3,000 parameters that you're looking at. Is that the filter process for you to collect all of the data -- like I drive Rivian. I come home. When does that data get downloaded? And how does that work? And how does that play into your advantage?

Yes. So it has a whole trigger system. So initially, when we launched the triggers were relatively simplistic. So it would simply be on things like gear changes or indicators. We now have a much more sophisticated trigger system that actually has machine learning in the loop as well. So we've found triggers that are interest -- we found events that are interesting, actually train a model to trigger on those and then ship that model to customer vehicles.

And in terms of how the uploads happen, we have a tiering method. So some triggers are considered critical. They'll actually go up by LTE. Other triggers are less critical. They'll go up via WiFi and then LTE only if we have to. And we can also push triggers dynamically. So we can actually -- you can sit at your desk, say, write a new trigger, push it and then you immediately get -- start getting that data back. So it doesn't have to wait for the OTA cycle to get the new triggers to the fleet.

And then how do you think about that sort of real-world data relative to the simulation data and the simulation process what is the efficacy of one versus the other? And I would imagine it's a spectrum that everybody is operating across where you're either more SIM heavy or you're more real world heavy. You guys are falling more and more on the real-world heavy side? Or -- how should I think about that?

Yes. I mean for me, there's no real replacement for real-world data. I think simulation can augment, but the real world always has the most interesting edge cases you can't really come up with. Sitting you can't really imagine them and they're often -- it often isn't the one thing. It's actually a -- it's a combination of multiple long-tail events that all sort of congregate into a single scenario. And so those things are very difficult to even think up and then actually quite difficult to simulate as well, -- so for us, it's really -- the real-world data is the kind of foundation.

Now we do have abilities to resimulate those real scenarios. And so that's something we believe in. But for us, we have a huge data hose essentially of data, and that's what we think is the right way to do these validations.

kind of the primary driver there. You've said RAP1 can easily be extended to solve other physical AI problems such as robotics. We've recently written a report talking about how we feel like the automotive industry, specifically in the U.S. is behind in terms of using their capabilities in the automotive space to enter into other spaces where they have similar applications. So can you talk about that opportunity for Rivian more broadly to focus on other end applications? Is that something that you take seriously now? Or is something that you see happening down the line because we've seen -- I think there's more than 15 Chinese OEMs, there's Hyundai, there's a couple of the European OEMs have all made an entry in some form or fashion into sort of the embodied AI space to leverage the capabilities that they already have from automotive.

Yes. I mean I think in Rivian's case, part of what we did is set up Mind Robotics, where as RJ was looking at the current robotic solutions, very specifically around manufacturing and the use case there for what we're thinking about that was part of the genesis then of thinking, how can we have extensibility into developing robotic solutions for manufacturing. Now it's not like a general purpose thing that we're trying to create. It's really focused on that specific area. But to a certain extent, we'd have to defer to the Mind team on where they're looking to head at this point. But part of the partnership with them there is going to be around the fact that we have an in-place manufacturing line and then being able to use that data as they think about developing their own solutions.

Now you can sort of deploy it in your own factories and then they get the data flywheel going very similar to the autonomous approach.

Exactly.

So what do you expect initial production capacity look like? And how do you prioritize RAP1 for Rivian before entertaining outside supply contracts?

Well, right now, the priority for RAP1 is going to be going into the R2. So for employees, the plan is to start rolling out RAP1. So our third-generation computer plus the LiDAR to go into the vehicle later this year for employees and then start rolling that out to customers next year.

Okay. Uber had invested. I think it was up to $1.25 billion through 2031 with $300 million at signing and then sort of a milestone based, you have $250 million targeted by the end of this year. what specifically triggers the $250 million? And then there is the 10,000 autonomous R2 purchase commitment? Is that a firm order commitment? Or is it contingent on some specific autonomy metrics?

Well, so we haven't given specific details in terms of what's going to unlock each of the funding milestones. The way you can think about it is going to be around achieving certain technical milestones out of James' team on our march towards getting to an L4 capable robotaxi and then the expansion of that service where we said the goal is by 2031 to have rolled out the service into 25 markets including at least 1 in Europe. So that's what would unlock each of those triggers.

But our expectation remains that we'll be able to unlock the second trigger for the $250 million investment here in the fourth quarter. In terms of the actual purchase of the initial 10,000 vehicles from Uber, clearly, we need to have gotten to an L4 capable service before they would want to purchase those vehicles. So we would need to make progress there.

Right. So -- and you haven't given any sense of timing on that publicly?

So we've said that the goal is to have the service available by late 2028.

And then once those purchases have been made those vehicles have been delivered, who owns and operates the fleet in 2028.

So Uber or their fleet partners will be the consumer purchasing the vehicle. So from a revenue perspective, we would then recognize the sale of those vehicles to Uber or their fleet partners. But then an important part is that we're going to get a software licensing fee associated with the fact that it will be Rivian's driver out driving every single day. And think of that as being tied to the utilization of the vehicle.

So is it sort of a dollar per mile.

Yes, we haven't given specific, but think of it around utilization as it's driving around on the mileage.

Okay. You also have personal Level 4 targeted for 2028. So which comes first?

So yes, we likely think the robotaxi comes first just because it will have more operational constraints on it. But the bulk of the work towards personal Level 4 will be the same work that goes into robotaxi. So one would be relatively close in time to the other. And then we do think that the North Star is really personal Level 4. As big as we think robotaxi will be, we think personal vehicle ownership will actually not go away. And the vast majority of autonomous miles will be on your personally owned vehicle longer term.

And so we -- I'm personally very excited about the vehicle that can drive me say, to Tahoe overnight or pick up my kids, and I think that's what a lot of people would want.

Does the Uber platform robotaxi deployment constrain your ability to monetize that personal Level 4 at all?

No.

Now it's kind of all additive in one.

Yes. And that's actually part of the benefit of the Uber partnership is by getting the $1.25 billion of capital, assuming we hit each of the milestones, that then enables us to accelerate the investment in R&D spend for James' team.

In the R2 ramp, moving on to discuss that a little bit. What are some of the key focus areas that you guys have been looking at internally from a supply chain management perspective that you're most focused on?

Yes. And RJ has talked about this a little bit on the conference circuit. But I'd say, first and foremost, it's really around ramping the supply chain and seeing that supply chain mature where as we think about it, it really is a symphony of needing to bring not only all of your Tier 1s up, but then the Tier 2s, Tier 3s and even potentially below that. Because, of course, at any given point in that chain, if there's any sort of a challenge, then that's going to be able to limit your production.

And so we've been very intentional from a ramp perspective of taking a crawl, walk, run approach going through our first design validation builds then our manufacturing validation builds, then having the first shift up and running. We're now hiring and ready to have the second shift get up and running as well as we head into the fourth quarter. So that's all milestones as we look to eventually ramp to get up to what we think will be our North Star of 4,000 units a week within the plant in Normal, which would get us up to roughly full utilization of the plant, but it's going to take time to go through that ramp process.

Right. Are there any specific areas, though, where you're particularly focused or particularly concerned or RJ is particularly concerned that he's talked about?

Not any specific area. I mean, as you would expect with a supply chain ramp, any given week, there can be one specific supplier focused on and then the next week, it could change in terms of who you're focused in on because it's really the supply chain right now that is constraining production within the plant, which RJ has talked about. So it's continuing to work through that maturation of the supply chain.

Yes. Can you talk a little bit about the BOM going from R1 to R2? And where are the biggest areas of cost savings that you're planning to see?

Yes. So we talked about this on the Q2 earnings call, but our expectation is that the bill of materials for the R2 is going to be about half of the bill of materials for the R1 and non-BOM costs will be even lower than that in terms of a decrease. Now some of that is making very intentional decisions on specific elements within the vehicle. So thinking about, for example, the suspension, where shifting from a fully active air suspension to a more traditional semi-active MacPherson strut configuration, that's a 70% decrease in the cost of the suspension. And there's element -- other elements of decisions where we made very intentional decisions to drive down that cost of the vehicle overall. So it's across multiple different vectors of the design within the vehicle.

Right. And then you've got the scale benefits that come obviously from your own cost side of things, but also from your supplier side of things. So can you talk a little bit about how your relationship with the Tier 1s and Tier 2s has changed as you guys have increased your ambitions in terms of -- from a volume perspective?

Yes. And that's -- it's really a virtuous flywheel where as we get more and more scale, you then have an ability to go back to your suppliers to push them more to get even more favorable pricing. So relative to back in 2020, 2021, as we were going through the process of sourcing for R1S, R1T and EDV, at that point, we were, relatively speaking, an unknown operator looking to go out and ramp an entirely new vehicle program or 3 new entirely new vehicle programs. And so we had to go in many cases, hat in hand to suppliers saying, "Hey, please work with us." Oh, by the way, there's a pandemic going on. So that was a very challenging operating environment relative to where we're sitting now with the R2, where as we were going through the sourcing process, working to get all of the components for the R2, we're now very much so a known entity within the supply chain and being able to have very fruitful conversations to be thinking about how do we drive down cost to get to the price point we want to get to for the R2.

So I think it's part of the benefit of the maturation of the business is that now as we scale up R2, being able to source appropriately for that. And then, of course, with our plant in Georgia, looking to add even more scale so that as the business continues to grow, it just gives you more and more leverage as you talk to your suppliers.

Yes. One of the things that has happened since you had the sort of idea and planning for the R2 is the industry has largely pulled back from BEVs. How do you think that impacts, a, demand; and then b, your supply and the cost that you have to pay for that supply because the battery industry is largely overcapacitized globally, specifically here in the U.S.

So how has that impacted the BOM? Obviously, it seems like it's probably a tailwind. And then also from a demand perspective, I think it's kind of 2 sides of the same coin, where you could argue that people are less open to battery electric vehicles. And so I think some of our investors have questions about demand because of that, you could also argue there's much less competition than you were probably previously expecting.

Yes. Well, and I think we would both argue that from an electric vehicle adoption perspective, we're now to the point where we're back to a more stabilized area, and we're going to see growth from there. I think part of it is -- and look, I own all electric vehicles. James, I think you own all electric vehicles as well. At least from my personal perspective, I'm never going to go back to owning anything other than an EV. From a performance perspective, from a total cost of ownership perspective, let alone looking at the price of gasoline now, it's just a better product.

And so I think from our perspective, part of what's been inhibiting the adoption of electric vehicles is bringing really high-quality choice to market. And you've heard RJ talk about this where we think one of the key drivers is going to be bringing more choice because from a form factor perspective, a lot of vehicles that have come out in the EV space have all looked relatively similar versus when we bring the R2 to market, it has a totally different look, a totally different field where it's an EV that is designed not just for being a good daily driver, but if you want to drive up to Tahoe and you want to go off roading or thinking about more of a traditional 2-box shape for what your vehicle looks like. And so I would say the biggest driver of more EV adoption for the industry at large, one is bringing more high-quality choice, and we think we're going to do that with the R2.

Now in terms of thinking about supply chain. Clearly, there's a lot of puts and takes in the market right now. Certainly, we are seeing and we talked about this on the Q2 call, we're seeing increasing commodity costs. We're seeing things like DRAM becoming an increased cost as well. So that can offset other places where you might be seeing some benefits.

But the cost of actual battery cells, has that come down versus what you were previously expecting? Or is that stable?

Lithium prices are, of course, going up too. Now we do hedge lithium, so we try to manage that price accordingly. But the raw inputs for batteries are also going up, too.

Yes. I don't know if anybody in the audience has any questions that they'd like to ask. We can pause for a second.

Yes, I'm just one upfront.

I have a question on EDV. Is there opportunity for that at any point in time? And in addition, is the extended range all-wheel variants, are those on track? And maybe when might we expect to see those roll out?

Do you want to take the first one?

Yes. So in terms of the stack, we are -- it is a general stack. So we can move sensors around, we change the vehicle configuration. We can support new vehicle types and different modalities as well. The way we would do that is, first of all, through the calibrations and then collecting a certain amount of like expert driving data on that platform that we can then use to fine-tune the model. So I think it's definitely -- we've built the stack to be flexible in that way and also to have the flexibility on the compute side so that we can support, for example, ACM, our second-generation computer and the third generation computer.

Now in terms of EDV specifically, we do support like active safety features on the current generation van, and we're always into -- I mean, we're actively investigating supporting future versions of the van that could have more capability in terms of autonomy.

Yes. And in terms of partnership with Amazon and rolling out the new variants, continuing to progress well and continuing to see good momentum with the partnership there.

On the -- one more in the back.

You mentioned developing your own silicon and in-house for future vehicles. Is that done entirely independently? Or is that in partnership with some fabless semiconductor providers or other licensing players?

So yes, so I may not be the person -- I'm not the hardware person, but -- but certainly, all the like TPU like pieces of that chip are all built in-house, and we own the IP on that. And then some of the other IP that's on the chip, we partner with others to provide. So that's the way we develop that. Yes.

All right. I think I'll hop back in. I think one of the most -- the biggest opportunities for Rivian to kind of surprise people and outperform is to do another deal that's very similar to Volkswagen. Can you discuss any of the color that you guys have been having in terms of an increasing penetration or other contracts or other OEM wins along those veins.

Yes. Well, nothing -- certainly nothing we've announced to date on clearly can't announce anything here. But I think as we think about the future end state of the industry, which we very much so believe is going to be electric software and AI defined and autonomous for a lot of traditional OEMs, at some point, they're going to need to figure out how to shift to a more software-defined vehicle. And that means at a very base level, having a zonal electrical architecture and then the ability to have software that you can update over the air to improve the vehicle over time.

I think proof -- the biggest proof point right now is around executing first and foremost with Volkswagen. So they have said that they expect to roll out their first vehicle with our joint zonal electrical architecture and software next year with the ID.1. And our focus then is if we execute with them, that's the best signposts as we think about potentially working with other OEMs over the medium to longer term.

And then James, on the autonomy side, I'll defer it over to you.

Yes. So the autonomy is separate from the JV with Volkswagen, and we did that very intentionally just because we have confidence in our own abilities to develop like a really good system and then potentially monetize that later. I think we're very open to kind of working with other OEMs. And I think as Chip mentioned, like every OEM is going to have to figure this out because I think the consumer bar or the expectation level of what types of autonomy available on the vehicles is going up and up actually pretty rapidly, I would say.

And so I think we'll be in an amazing position when we ship and demonstrate point-to-point in our customer vehicles that we have a technology that's really valuable and -- but other OEMs will need to have some version of.

Yes. And then from a demand perspective on the R2, obviously, the -- I think all of the initial launch expectations reviews were very positive. How do you think about that order book growing over time? I think one of the things that we're trying to get our heads around is just -- as you're ramping production, that order book should either be staying the same or growing more because you'll have word of mouth. Is there another advertising push that needs to be made as we go into 2027?

Yes. I mean as you think about what we're going to be doing, one is from a quality perspective, making sure that our best salespeople, which is our customers are very happy with their vehicles from a manufacturing ramp making sure we're ramping with quality. Then second is starting to drive more visibility in terms of the brand. So I would expect to see us becoming much more active thinking about our marketing efforts to drive visibility of Rivian because look, admittedly, we -- here we are in New York and James and I both live in the Bay Area where there's a much greater penetration and visibility for the brand.

But we do know that getting higher levels of visibility and folks in other parts of the country becoming more aware of the brand is going to be really key as we think about continuing to drive adoption of R2 over time.

What about the opportunity in Europe for you guys to expand there?

So Europe, we continue to expect that Europe is going to be part of our story over time. We haven't announced when our plans are to get over to Europe. And certainly, there's any number of dynamics that will determine when we'll eventually get there. We could potentially ship vehicles from Georgia, for example, or at some point, you need to do manufacturing over there. So we'll have to see first how the ramp is going here in the U.S. and what demand looks like. And then at some point, we'll make that decision for when we want to head to Europe.

Yes. And then maybe if you could just -- I know you've laid it out publicly for investors, but if you could just for our audience discuss your capacity plans?

Yes. So from a capacity perspective, right now, our plant in normal has an annualized run rate capacity of 215,000 units. Of that, we've said roughly 155,000 to 160,000 units would be for R2. And then the goal -- we've already broken ground and are currently working on the Stamping Press Foundation and working on the foundations in Georgia. So as we progress on that, that will eventually become a 300,000-unit facility in terms of annualized production in Georgia to take us to over 0.5 million of productive capacity between the 2 plants.

Georgia, however, is going to be more focused on our midsized vehicle platform. So R2 and associated variants, R3 as well as then the robotaxi with Uber.

Okay. Any more questions from the audience? No. All right. I think that's enough. You guys have been great. Thank you very much. Really appreciate it.

Thanks.

Thank you.

All right, everybody. We'll get started with the next session here. Very excited to be joined by Founder and CEO of Rivian, RJ Scaringe. So thank you for joining us today. Just a quick disclaimer before we get started here. So for important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley representative.

Also, just a quick plug. There's 3 Rivians outside. If you haven't gone up and test drove them already, I strongly recommend you do so, find some time between your meetings to do it. It's worth the experience.

But maybe just to get us started here, it's a very important year for Rivian, a lot is happening. So I would love for you to just kick us off, RJ, with maybe a few minutes of opening remarks in terms of what you're focused on and kind of where investors should be spending their time.

Well, first, thanks for having me. It's great to be here. For us, this is like an inflection point year. We launched our R1 product, this was in 2021, now almost 5 years ago. And that was intended to be the handshake with the world. It was our first vehicle. It was the first exposure to the brand. It helped to establish who we were. And off the back of that, built incredible brand loyalty, a lot of enthusiasm around the company.

But ultimately, the price point of R1 meant that it was not a mass market product. And so over the last few years, we've been working towards, and I've certainly been talking about it, R2 is really the mass market vehicle and the vehicle that takes this positive brand sentiment, positive enthusiasm what we're creating and puts it to scale.

And so that inflection point is starting as we speak. So we launched R2 just a couple of months ago. The feedback around the vehicle has been just outstanding. That's everything from journalists and influencers all the way through to customers and just the level of engagement and I'd say the product market fit is exactly what we'd hoped for. And so now as we think about what the rest of this year looks like and then into next year, of course, our focus is on ramping up production. That's not just our production plant, but making sure our suppliers are capable of supporting the ramp-up and delivering a lot of vehicles to excited customers.

And so alongside of that, we've been really focused on our technology. And so within that, you see the technology that underpins our software and electronics. We did a deal in late 2024 with Volkswagen for a $5.8 billion licensing arrangement around our software and electronics to support a host of brands under the Volkswagen portfolio. And then recently, we did a deal around our Autonomy platform with Uber to create a robotaxi version of R2. But ultimately, all those technologies also manifest into the consumer products and create an even more differentiated experience. And so a huge amount of effort and focus is continuing to go into that area of the business as well. But yes, we're -- as you said, this is an important time, really excited time -- exciting time. And we -- I'm loving seeing more and more R2s on the road.

That's great. And maybe just in terms of that customer demand and what you're seeing in terms of reservation conversion, I think on the second quarter call, you mentioned that you were seeing, I think, a larger percentage of people deciding to go with the launch edition of the vehicle. Just maybe provide some updated comments. Has that changed at all over the last few weeks as there's been more people driving the vehicle? What's that conversion been from test driving the vehicle to putting in a reservation? Just any additional color you can provide in terms of how the demand profile has trended over the last few weeks would be great.

Yes. I guess just as a reference, for everyone here, the way we launched R1 is we launched out of the gate with a lot of different build combinations. So essentially, everything you could buy was available on day 1. And we learned a lot from that. It was a very complex way to launch. We introduced massive complexity into the build process and our supply chains from the beginning.

So with that as a reference, in contrast in R2, we launched with a very narrow set of build combinations and in fact, one trim, which we called our launch edition. And that launch edition is our most expensive trim on R1 (sic) [ R2. ] So it's a $57,000 starting price vehicle, very limited paint and color options. And the whole objective was to simplify the initial ramp-up. But of course, alongside of that, we had to make assumptions around how many people would convert to go to the most expensive version of R2. And that conversion has been higher for those -- this early batch than what we expected, which is a really good sign. So folks that maybe thought they were going to buy a $50,000 version of R2 said, fine, I'll take the $57,000 or $58,000 version. And I think it's just reflective of the excitement around the product that people are willing to be more flexible in the version that they were taking far more than we expected.

Now as we look at the next 12 months, we'll introduce additional trims. And so we have a premium trim that takes price down into the mid-50s. We have some base -- like we call standard trims that are in the 40s. But ultimately, within the next year, we'll have a starting price of $45,000 all the way up to the maximum price, which is just over $57,000. And that's -- that portfolio will support an ASP that probably sits somewhere in the middle.

But the excitement around all the different trims, we've had a chance to expose media to the different trims and allow them to drive that. I'd say we've done a really good job. Of course, I'm biased, but we've done a really good job of creating a strong value proposition across each trim configuration and each version of the vehicle that you're getting a lot of value for money.

Yes. And does that change the cadence for some of the lower-cost variants to come to market in terms of the amount of people that are deciding to go with the launch? Or are you still tracking for performance launch next year? And just curious if there's been anything shifting in timing just given the demand on the launch edition.

I mean it's a good question. There's obviously tweaks to exactly when the different variants come in that we're making. But broadly speaking, it's about the same.

Okay. Great. And then in terms of the customer base, right, I think you've talked about wanting to attract a broader customer set to the R2, any data points you can provide in terms of as you look at the reservation book, what percentage of those people already own an R1? What percentage of those people are -- just purely don't own an EV yet, but are very interested in the R2? Just trying to get a better picture of the demographic of and the diversification of that order book or reservation book.

One of the goals of R2 was to make sure that we're establishing a really wide net to get as many people excited about the brand and the product as possible. And when we think of R2, because of its price point and because of its size, I think it's natural and not sort of expected that a lot of folks tend to say, "Well, how does this get cross-shopped with the Model Y?" Price points are very similar. The R2 is slightly smaller than Model Y in terms of -- while it looks bigger because of its form factor and the shape, it's slightly shorter and slightly narrower. And so naturally, that's a question asked.

But the reality is the opportunity is the vast majority of folks who aren't buying Model Y. So that's folks that are maybe in a Toyota RAV4 or 4Runner, these are -- in Ford Bronco or Honda CR-V or Audi Q5. So there's like a broad spectrum of vehicles that exist in this midsized SUV crossover segment. It's the most popular segment in the United States and makes up about half the demand in the United States. And of course, the average price of new car being about $50,000 in the United States, from a price point of view, we're right in the sweet spot or the broad thick part of the demand curve.

And so that was the goal to bring, call it, like regular consumers in, not just early adopters, not just tech adopters, but our regular customers thinking about buying a vehicle to take their family to go on trips or to take the kids to school or to drive to the office. And so that's what we're seeing. We're seeing a very broad spectrum of consumers coming in.

What I'm really excited about is a lot of the folks that are -- have been activated by R2, this is their first time considering an electric vehicle. And so that represents a really interesting opportunity where for the test drive, in many cases, the test drive is the first time someone has driven an EV. And of course, if you haven't driven an EV before and you jump into an R2, like the experience is, it's somewhat otherworldly, like the vehicles accelerates really quickly. It's incredibly quiet. It's refined. The software is very snappy and feels like a piece of consumer electronics.

And so as a result, the conversion from test drives has been quite positive. And you've probably heard me say this before, but I think one of the most important leading indicators of future demand is the volume of test drives because the conversion ratios are quite high. That's such an important factor. It's the thing I track most closely is how are we doing in getting test drives because that will ultimately lead to continuing to build top of funnel, mid-funnel and all the way through purchase.

Where are you today on that? Is that a number you guys have talked about in terms of number of test drives for R2? I know on the 2Q call...

We have, I'd say, maybe without being -- we've given actually specific numbers. But the way to think about it is that R2 because it's such a broad appeal product, it's a mass market product, it's significantly increased the number of test drives. It's more than doubled where we were before. But we continue to see that grow over time. We think that will be a growing metric for us and something we continue to track.

That's great. And so it seems like it's becoming more -- less of a demand question and more of a ramp execution, supply chain story at this point in terms of volume growth as we think about the remainder of this year and into 2027. Can you just talk -- you're a few weeks into customer deliveries. Just talk us through where the main pain points have been. Maybe don't call specific people out, but just curious if there's been any particular areas of pain that you've had to solve and what the process has been to do that?

Yes. So if we think about getting to much higher volumes, there's the demand side and there's the production side or supply side. And I think naturally, when we think of producing R2s, our minds go to thinking about the Rivian production plant. And that's -- like that's the question I get asked most often, like "How are things in the plant?" And that's very important that we ramp that effectively. But that isn't actually what's constraining our ramp-up. Our ramp-up is gated by or throttled by the rate at which we ramp up our suppliers.

And to state the obvious, I think the difference is often that, that comparison is drawn because when we saw Tesla go through this similar inflection point with Model 3, in their case, their constraint was initially in the plant. In our case, our plant is out pretty far in front of where we are from a supply base point of view. And so getting the hundreds of suppliers to all ramp at the same rate, and we can only ramp as fast as our slowest ramping supplier. That's been our focus.

And so now saying all that, that's not unexpected. It's not like this is a surprise. We -- if you'd ask me this question a year ago, I would have said, I think the throttle for a ramp and what defines the ramp curve will be our suppliers, and that is, in fact, the case. And this is something we've planned around for a while. And now, of course, the work we're doing is to make sure that all of our suppliers continue to ramp up as planned and continue to achieve those targets because if a supplier isn't ramping at the expected rate, everything else has to be adjusted down. So all those other -- we wouldn't want 99.9% of our suppliers ramping and then one not because we'd be building up insane levels of inventory of the -- all the parts we have and be constrained by the one part.

So think of it as this big symphony where everything has to move together. And it's often not the Tier 1s. The supply chain has been unstable enough for quite some time where you just have a lot of uncertainty within the supply chain, but it's the Tier 2s and Tier 3s that can be the most challenging. And so if you -- again, just to decompose this or deconstruct this, if you think of our supply base as having several hundred direct suppliers to Rivian, each of those suppliers, it depends on the supplier, but on average, say they have 10 suppliers. Those are Tier 2s. And each of those Tier 2s probably have 5 to 10 suppliers as well. So we're talking about many thousands of companies that are responsible for making the components to ultimately make their way into a vehicle. And any one of the suppliers in that long chain can cause a disruption.

Now of course, as you move further and further upstream to the supply chain, ultimately getting back to raw material to like mines, there's lots of inventory in the system. So there's more buffer, if you will. So it's like if there's an issue at a mine, that's not something that we immediately feel the next day. But if there's an issue at a mine, we want to know and make sure we have other sources of supply that then feed the -- often like a Tier 4 or Tier 5 that feed the Tier 4 that feeds the Tier 3, that then feeds the Tier 2, that then Tier 1 and then us.

And so that's the nature of this. And I think because trade has been so dynamic and the trade relationships, especially at the Tier 2, Tier 3 level has been so dynamic, it's also just been something we've had to really manage carefully.

And would you say it's gone so far as expected? Is it faster, slower? I'm just curious in terms of if you were to go back 6 months and say where you want it to be relative to where you are today.

Well, we've provided guidance for the full year that reflects our expectations. We've made a slight tweak to that over the course of this year. But so I'd say guidance-wise, we've been pretty consistent. We, of course, are working to go as fast as possible. So if there are ways to like go faster and accelerate, we're pushing on those, and we know what those levers are. We know what the slowest ramping supplier, we know exactly who they are. We know what the issues are. We have people on site. So we're trying to go as fast as we can because we see the backdrop of so many customers excited to get their cars.

Right. And then the other piece, you talked about adding a second shift by the end of this month, I guess. Is that still on track? Where does that process? Where do you stand in terms of getting that second?

Yes. Second shift is on track. Again, that needs to be coordinated, bringing a second shift on to then not be busy because you don't have parts is also a challenge. So we want to coordinate that really tightly with the ramp-up of the supply chain. But that's been choreographed really nicely. We have a great training program for bringing on new resources, both at the team member level, but importantly, at the team leader and group leader level within the plant.

And when you think of like plant floor leadership, this really comes down to the group leaders and the team leaders. This is like -- this is the engine room, so to speak, of the plant. And so we've invested a lot of time into that function and into those roles, into those leadership roles within our operations.

Okay. Great. And then I know the other area that's you're super passionate about is autonomy and the path to autonomous driving. And you guys have communicated a desire to get to point-to-point eyes-on driving by the end of the year. Maybe just talk through where you are. Obviously, we're getting closer to the end of the year. What's kind of left to be done to get that rolled out to customer vehicles?

Well, we have -- we're going to announce the specific date, but we're going to have an Autonomy and AI Day that we have near the end of this year, similar to what we did last year. And the goal of that is to both provide an update on all the progress that's been made over the course of the year. Of course, since our first Autonomy and AI Day, we had some announcements, one of which was a partnership with Uber to deploy a robotaxi version of R2 in late 2028. So we want to be able to talk about the progress we're making on the technology and then also use that as a -- think of it as like an anchor point to then roll out some of these new features that we're really excited about.

And I'd say that on the staircase to Level 4 and the vehicles being capable of delivering robotaxi capabilities or full autonomy, a really important step is point-to-point. And that's -- and just to make sure everybody is clear on what I mean -- what we both mean when we say point-to-point, this is where you get into the vehicle, you type the address of your intended destination into the screen and say go. And the car does everything from there forward. So it pulls the car out of a garage or out of a parking spot, drives to the final destination, finds a parking spot at the final destination.

And so that's -- we think that's a big unlock in terms of how people perceive autonomy. It's by no means the final unlock. This is still, call it, early days for what ultimately is going to come, we think, in the next few years where cars -- increasingly, it's an expectation that the car just completely drives itself, meaning it can drive itself empty, it could drive it with you in the back seat. But on the journey to that, this point-to-point where it's, as you said, supervised. So your hands are off the wheel, but your eyes are still on the road most of the time. When I say most of the time, I mean, you can look down, be distracted for a bit, but you can't take an nap. You can't watch a movie.

The obvious next step beyond that is then eyes off. So you've heard me probably talk about hands off, eyes off, but the ability to be completely disengaged from the driving. And then, of course, the step beyond that is Level 4, which is you don't even need to be in the car. And so we -- that road map is something we see really like playing out meaningfully over the next 2 to 3 years with 2028 being a really important year for us where we start to deploy the Level 4 version, which is this robotaxi version in a customer application where you actually have it working, driving cars, earning revenue while it's driving.

Yes. And you raised the point about the point-to-point being like the real unlock for customers where it's like a -- there's an added value that comes with that versus just being able to keep it on the same lane or on the highway. That is a very different functionality and technology.

And I'm curious, as you've engaged with customers, to what extent that is a selling point? Or are they waiting for that? I'm assuming people who are already in an EV to the extent they're in a Tesla, know what it's like to have point-to-point. But someone coming from an internal combustion engine vehicle may have never experienced it. They might not know what they're "missing" at this point. So I'm just curious like what the feedback has been in terms of customers wanting to see that on the vehicle before they're ready to make that purchase decision for the R2?

I love this question. I think it's important and all of you in the room spend time thinking about this segment. This is a very unique moment in time where in our own bubbles, it may seem like everyone is thinking about and knows about autonomy and the nuances of it and the difference between Level 2 and Level 3 and hands off, eyes off and hands off, eyes on and all these things that I can say quickly in here and we know what I'm talking about. But the number of actual customers that today know about that is actually quite narrow.

And so as you said, folks that are in a Tesla because of exposure to FSD, they have more awareness of this. But the vast majority of customers buying new cars in the United States today are not making those purchase decisions around autonomy. That's the way it is today. Now here's what's, I think, the important giant like part that we need to underline in bold face is as autonomy becomes more capable, this becomes a very important purchase criteria because you start to get your time back.

And so a core thesis that we hold to be true and believe in is that we are going to see an inflection point where this goes from a small subset of the buying public that's early adopters, maybe buyers that are buying EVs today to everybody in the United States, everybody in the market is going to want to have the ability to drive to work without having to pay attention to the road. We think that plays out over the next handful of years.

And so for that to happen, there's a lot of education, test drives are going to play a key part in this. But we do think it's something that's going to dramatically shift market shares. It's going to dramatically shift the topology of essentially who's providing the technology, who's providing the vehicles. It's going to lead to a very different looking auto industry as we think about the early 2030s and into the mid-2030s.

But all that said, right now, the focus is get that technology right. And then, of course, with the recognition that the early adopters know a lot of these details, care a lot about it and then make sure that customers who are buying Rivians and buying cars for reasons that are not autonomy, know these features exist because word of mouth is going to be powerful here.

Yes, absolutely. And I would say, would you say that, that is from a pricing standpoint and a willingness to pay for that functionality, where -- what's the balance, right? Because at some point, if it becomes a requirement for this technology to be on a vehicle for it to be deemed even worthy of and being on a consumer's mind, one could argue that pricing at some point might have to come down. But I'm just curious, as you think about the next kind of 3 to 5 years, where your pricing is today and what the willingness to pay is. And maybe it's a much more complex conversation because you get your time back and you have to factor that in, some productivity gains. How do you think about kind of the pricing algorithm for your technology...

Well, I mean, ultimately, it's going to be market-based. It's -- and so I mean, I'll tell you my point of view, but the two bookends of how one might answer that question is, on one end, if you -- is the view that autonomy -- high levels of autonomy is going to be -- it's a scarce technology. And so there's only a few providers of that. And the other end is it's commoditized. It -- probably not too surprising. I believe and Rivian's strategy is built around this belief is that we think it's going to be a fairly scarce resource, and we think it's a resource that's valuable to own. And so we're spending a lot of money and time developing that.

I think the number of companies that will likely have it is certainly less than the number of fingers on hand. It's like -- it's a small number of companies that have a legitimate and very plausible path to get to these levels, meaning you have robust data flywheel, access to low-cost inference for running large models in the vehicle, really highly robust perception platform, camera heavy, but in our case, camera heavy plus radar and LiDAR, both -- all three of those perception platforms being low cost. And so we think there's a narrow set of companies that fall into that category.

And so as a result, we do believe there's going to be persistent pricing power that exists for customers that want to pay for this. And I think, again, this is my point of view, Rivian's point of view. I think every manufacturer to maintain market share as we think about the next 5 to 10 years will need to have software-defined architecture, so the vehicle needs to be software-defined, which today is really just Tesla and Rivian. When I say that, meaning not a collection of third-party ECUs that run independent OS islands, but rather a consolidated set of ECUs that run a common OS. And so today, two companies do that outside of China, Tesla and Rivian. And on top of that, a self-driving platform that's capable of delivering very high functions. So ultimately, Level 4 functions.

And so I think if you don't have those two capabilities, you will lose market share. And so I think every -- that's a belief. So on the back of that belief, I think every manufacturer has three choices. First -- number one is three paths you can take. You can ignore it. You can say, I don't believe that those things matter for market share. We're not going to focus on that. You can build it, you can develop it yourself, which requires very capable teams around software, AI, electronics, model building. Or you can buy it, and there's going to be a fairly, I think, limited number of folks you can buy that from.

But what I just described in that framework is exactly the framework that led to our deal with Volkswagen, where Volkswagen decided on the software OS, the first of those two big buckets I talked about, to say we're going to work with a company that's proven that can do it. I think there's lots of opportunities in the software world to do that. On the software platform, I think there's going to be a similar set of big opportunities for autonomy. And as a result, for us, we can monetize the technology platforms that we're building either through market share or we can monetize it through providing it to other OEMs or we can do both, which is, of course, you can see thus far, our strategy is to pursue both, both pursue growing market share, growing revenue through vehicle sales, but also recognizing that these technologies can be sold independently to manufacturers that have decided to not build it themselves and have decided to act as opposed to ignore these big megatrends.

Yes, absolutely. And I would say -- or ask like when you think about the robotaxi partnership with Uber, you think about -- you talked a little bit about tech licensing. But if you go the other direction, you talk about your partnership with Uber, to what extent is that program important for you to achieve or at least expedite your path to eyes-off point-to-point in terms of the data capture you get, in terms of the experience and having these vehicles on the road and the driverless unsupervised capacity, is that a requirement to get to eyes-off for personal vehicles or just expedite things for you?

Well, it's a great question. It's probably helpful for me to just maybe ground us all to describe how we've approached self-driving. And so when you think about self-driving and the technology necessary to deliver on this, all of us recognize that this has been something that's been worked on for a while now since, call it, I don't know, 2012, 2013 when you started to see a lot of concerted efforts around this. And it would be easy but wrong to assume that the work that was happening in 2013 was highly related to the work that would be happening today.

And the reason I say it'd be easy but wrong is, of course, the output, the application is the same. It's the vehicle drives itself. The difference is we had a significant shift in how we approach building these systems. And so the way self-driving was approached up until just a couple of years ago was a vehicle would have a perception platform on board. So it's cameras at a minimum, but at a maximum cameras plus radar plus LiDAR. That perception platform would identify and classify objects. So that's a car, that's a bike, that's a tree and then it would associate vectors to those objects. So acceleration, velocity to all those different objects.

And so what would then happen is that set of classified vector-associated objects would be handed to a planner. And what the planner was, is it was -- think of it as a large code base that was written by humans to then describe driving. So it would describe what to do with all these various actors in the system.

And the limitations of that were obvious. It was very much specific to the vehicle application. It was very much specific to the geography and the area that it was operating, and it was quite fragile. So if you took that platform and put on a different vehicle, it required a lot of work to redescribe how to drive the car. If you took that platform, moved it to a different geography, a lot of work to describe how to behave in that environment.

Of course, what happened, and we witnessed this in the most fantastical way imaginable with LLMs, is the introduction of transformer-based encoding, the idea of building large multidimensional neural networks and the concept, which is not so common for us to talk about with foundation models and this idea that a much more complex multidimensional model can capture very complex bodies of human knowledge in ways that are not this sort of rules-based approach that we've tried to do in the past.

So of course, in the LLM world, you saw -- we saw things like Alexa and Siri in 2017, 2018, which could maybe tell you a joke, maybe tell you the time of day. Now advance to LLMs, which are like there's absolutely 0 relationship between what was there in 2017, 2018 in LLM, when an LLM can give you relationship advice. It can write a book report for you. It can diagnose a red spot on your arm. It can do all these things that, of course, you would never even think to ask Alexa a decade ago. And so I call out the Alexa versus LLM because I think it's illustrative because that's the same dynamic of what existed for autonomy pre-transformer.

And so all the work that happened from the 2010s up through early 2020s, was what I'll call autonomy 1.0. We're now in autonomy 2.0 where we're not trying to write down in a set of code like how to drive, but we're collecting very large pools of data. And so to do that, you need a big data flywheel. Of course, 2 companies that have effective data machines are Tesla and Rivian. Tesla is, of course, the biggest because its fleet is by far the largest.

That -- all that data gets ingested. We use that to pretrain a big model. In our case, we call that LDM, large driving model. But there's only one model. We don't build an LDM for Level 2++ and an LDM for Level 4. It's one big model. And in our case, we're feeding that model with a really optimized set of perception. So we have a set of cameras that's been designed not just to drive the vehicle, but it's designed to learn as fast as possible. So we've placed them and [ designed ] them to absorb all this information. We've -- alongside those, we've put those with radars. We have 5 radars in the vehicle. And then soon, we're going to launch a LiDAR and the LiDAR turns every single vehicle that has one that's deployed into part of this really large ground truth data collection fleet for training our foundation model.

And anyways, your question was -- coming back to that, your question was, does point-to-point is that a necessary step or like an accidental step, if you will, to get to Level 4. It's just part of the continuum. And so as we think about it, LDM just needs to keep getting better.

And this is the part that, I think, perhaps the most confusing for customers is if you're in a Level 2++ vehicle, the way that it feels is actually should be nearly identical to a Level 4. The difference is just how confident we are in the system such that we allow you as the operator to cease to be part of the operational control of the vehicle. So Level 2++ and Level 4, the difference is you don't have to be in the car, you don't have to pay attention. But the car is going to be doing for the vast, vast, vast majority of the time, an identical set of tasks, but it's effectively a look at how capable we are handling the corner cases and then therefore, managing those in a safe way.

And the first step to that is allowing you to take your eyes off the road to say, hey, look, the car is very capable such you don't even need to pay attention, so it's no longer supervised. Then the next is you don't even need to be in the car because if it does run into -- it's not going to run into an issue where it needs to ask you to retake control.

All of that said, one of the things I do think is worth calling out, which gets back to my point on the shift from early adoption to mass adoption, if you're a tech early adopter, if you're one of us, you can live in this world where it's supervised, you have this partial role. For a lot of people, that's like confusing. But soon, that will just be like "Remember when, like we used to have to sit in the car and pay attention while it was driving?"

That would be like this little brief -- we should all like enjoy this while it's happening because we get to tell people we used to do this. But this is going to be like a very brief moment, a speck in time before we get to a world where it's just like, "Of course, the car drives itself. Are you kidding me, like you guys had to like operate the car?" It will seem so foreign to -- like my kids will look at that and just be blown away that, that was the way it was.

Are they going to have a license, do you think? What's your prediction?

I think they will because I'm influencing them because I like to drive. But I don't think they'll need to. In 10 years, I don't think they'll need to. I think it will be -- it will sort of be like driving a manual transmission. You don't need to drive one, but people do it because...

It's cool to do.

It's like...

A novelty.

It's a novelty, yes. But anyway, a long way of saying point-to-point is a key step on that journey, but it's not the end step. It's just -- it's one of the steps.

That's great. Well, I think we're out of time here. So RJ, I really appreciate you being here. If you have not driven the car yet, go upstairs, take a test drive, it's worth it. I appreciate it. Thank you.

Okay. Great. Thank you, everybody, for joining us. My name is Mark Delaney, and I cover Rivian for Goldman Sachs. I'm very pleased to have with me today James Philbin, SVP of Autonomy and AI; Vidya Rajagopalan, SVP of Electrical Hardware; and Chip Newcom, the VP of Investor Relations. Thank you all for coming.

Great to be here.

Great to be here.

Well, one of the really exciting things out of Rivian recently has been R2, and the company recently launched that product, which, as many of you know, is a 2-row SUV at a more affordable price point than the 3-row and more premium R1 line that's been out previously.

Chip, maybe I can start with you for this one. Can you share more on what Rivian is seeing with respect to R2 demand and feedback now that the vehicle has launched?

Well, first, Mark, thanks for hosting us today and great to speak with many of the investors here. Hopefully, many out there have already gotten a chance to go out and demo drive an R2. But still early days in terms of what we're seeing on the R2. But most importantly, from a media perspective, the early feedback we've gotten, it's been absolutely phenomenal. I mean, this weekend, The New York Times posted their review. We've seen many number of comparisons against both, ICE vehicles, EVs and across the board, it's been really great feedback.

So I think from our perspective, as a company, it's showing that, from a product market a bit, we're absolutely hitting the right points for what the consumer is going to be looking for. And now of course, it's all about execution and continuing to ramp our supply chain, continuing to ramp within our plan to normal but very excited about the early feedback we've seen.

One of the metrics the company shared on the last earnings call was that conversions from the preorders were actually tracking ahead of what you had anticipated. I don't know if you can share more on how you're measuring that. Any context you can give on that front?

Yes. I mean I think it comes down to -- so for a little bit of context for everyone out there, so we've been taking $100 refundable deposits for our potential customers if they'd like to purchase an R2. And then we've been inviting folks in waves to then be able to come in to our systems, configure their vehicle and put in their intent to order and purchase, and then starting to deliver against that.

And so as we've been inviting those waves in, seeing increasingly a very good conversion rate of folks actually saying, "Okay, yes, I do want to buy the vehicle." And it's interesting, as we're now introducing new trims, for example, Coastal Cloud, we just launched, seeing, as we introduce each of those new waves, more interest coming through from customers, so continuing to see good progress there.

That's great. And in order to meet the demand, the company has been scaling up production. So you'd already started a single shift in Normal to make the R2. The plan was to add a second by the end of 3Q, so this current quarter or so, just a few weeks left. How is that ramp progressing?

Yes, still planning to add a second shift here in the third quarter, so working through the ramp process, as you would expect.

Any key bottlenecks for the company to solve?

Another thing that we've talked about, continuing to work through as with any ramp, working through the process of all of our suppliers, we can only move as fast as the slowest supplier in our supply chain, so working through that process.

Great. Well, it's a helpful overview of the R2. We are very privileged to have James and Vidya from the tech team. I know you have a lot of various projects you're working on and a great opportunity for us to dive into some of those. Well, one of the key themes, of course, at this conference is AI and applicable for Rivian in a few ways but including some of the autonomy products. James, maybe I can direct this one to you. Where does the company stand with its plan to have supervised point-to-point driving available later this year and then eyes-off for 2027?

Yes. So I think the team has been very hard at work on point-to-point recently, so scaling up the model, scaling up the data and the validation of that model and then actually putting the vehicles on the road in a sort of test framework so that we can get feedback and feed it back in. So I think we're seeing very good progress. Actually, we recently gave demos to the Uber CEO and CFO. I think one of those folks posted about it on Twitter. So yes, great feedback, and I'm very excited to get that feature out towards the end of the year.

And what are the biggest technical challenges remaining?

I think mostly around sort of dialing in a good humanistic behavior in all cases, so specifically around making sure the model is not too conservative, that it's sort of tracking speed as you would expect. I think one thing we've seen is actually very good behavior around relatively complex scenarios like construction or narrow negotiation, so very excited about the progress that's being made. Obviously, the validation needs to happen so that we can make sure it's safe before we ship it.

And similar as you think about moving from the point-to-point supervised to the eyes-off in 2027, so same sorts of things you're working through?

Yes, it's essentially the same system. So just imagine the continued development of this end-to-end model, where we gain confidence, we can validate on more long tail cases. We can scale up the number of miles that, that system sees. And eventually, you get to a confidence level together with the system architecture that can allow you to take that eyes off.

One of the things we've heard so far today at the conference, more broadly, has been about tightness in compute capacity to train AI. And as you think about trying to hit some of these milestones, I mean, are you able to get enough supply to train your AI?

Yes. So thankfully, we have obviously very close relationship with Amazon, one of our biggest investors. And through AWS, we have a great supply of GPUs that at least for the next 6 to 9 months is going to support us well. Past that, of course, we'll need to see and I'm very interested to see how this market develops in terms of compute.

And just maybe for any of you, and Chip, feel free to chime in on these as well. But the company has indicated that Autonomy+ adoption is trending positively. What have you learned so far about customer willingness to pay for this product?

Do you want to?

Yes. So I think, yes, generally tracking actually better than we expected. I think one thing I'd say is that the product is very kind of sticky in the sense that once customers have these features, they really struggle to go back. And I've actually seen this myself. Testing point-to-point is -- I really hate going back to vehicles now that don't have that feature set, even though it's at an early stage.

So I think in general, those products are very sticky. People really love to use them. I think, over time, it's going to become increasingly table stakes for any new vehicle to have these advanced autonomy capabilities. So that's kind of what we see. And yes, all of that is driving, obviously, demand for R2s and R1s and also Autonomy+.

As you're engaged with customers around people purchasing R1, R2, what are they saying around whether or not those buying decisions are being influenced by what Rivian offers on this autonomy software side? And is that leading to incremental sales or maybe even any lost sales as you think about where you stand relative to competition?

So I think you typically see 2 maybe cohorts of customers. One cohort comes from the traditional ICE world, where obviously, the autonomy levels are much, much lower. And there, I think people are looking much more around the size of R2, its capability, how fun it is to drive, the stance you have in it. And that's where I think, yes, R2 is really shining.

And then on the EV side, where people may be coming from Tesla, which have higher expectations, I think Universal Hands-Free, which is what we launched last year, is going quite a long way with those customers, but then the final set of that cohort is really looking for a point-to-point type feature. So we're hoping we can really knock it out of the park with both cohorts at the end of the year.

Well, that software doesn't need to run on hardware, so Vidya, maybe I can go to you for this next one. And at the AI Day back in December, you presented and were nice to take some of our questions. But you outlined a number of initiatives, including some custom silicon, the RAP1 processor. Maybe talk about why you made that choice and how important is custom silicon to Rivian's long-term competitive position in technology.

Absolutely. I've always said there are 3 reasons we've built our own silicon, and they're performance, cost and velocity. I'll start with the cost. It's pretty obvious. When you buy merchant silicon, there's a lot of margin stacking going on. And in this day and age, AI silicon commands quite a bit of premium. And if I can shave off a big piece of it, that really helps us offer more at the same price point. So that's the number one reason.

Number two reason is performance and in particular, because we are a vertically integrated company, we actually understand the vehicle top to bottom. We understand the application really well, and we're able to actually design a piece of silicon that is tailored towards autonomy, tailored towards physical AI. It is not a piece of silicon that was designed for a data center inference product that then migrated over here. It's really purpose built for this purpose, and we believe that you can actually get the best performance as well as the best cost because we don't have super fluid stuff that comes from a previous vestigial piece of hardware that's sitting around there.

And at the same time, I've spent a long, long time prior to this in chip companies and often chip companies designed to benchmarks or sample applications or open source models but actually having the real model we run, having that and co-developing the hardware along with it helps you really design the hardware optimally. So you get really good performance for it.

And then finally, the most important -- another very important and most important part is the velocity. As James' team and my team are co-located and when we find issues, we can work through them very quickly, which may seem like a trivial thing until you actually work at a company and work through that process. If you have -- buy silicon from anybody else, you find a problem, you file a ticket. Then their solutions or support team looks at it. They're not able to solve it. They escalate it to the engineering team. And then it takes -- usually, it's a matter of weeks before it gets taken care of.

And when we're working with our teams, usually, the teams walk over to each other, discuss it and gets fixed much sooner. And that's important because that's a virtuous cycle. You can move and iterate much faster. And all 3 of these over generations really help you build them out, and that's what we're after.

All right. As you think about the challenges that come with developing that silicon and some of the milestones you still need to maybe solve for the Gen 3 hardware, what does that mean for time lines? And is the Gen 3 hardware stack still on track for late 2026?

So building silicon -- and that's a very good point -- is not for the faint of heart. It comes with a lot of risk because a simple matter of fact is the wafer, single wafer cost -- to tape out costs a lot of money, and then it's very expensive. And then when you put it in the oven, it comes back after 4 months, and so the cost of a mistake is very high.

So you have to spend a lot of time doing validation pretape-out, which I'm happy to say we're at the tail end of this. We've had silicon in-house for more than 1.5 years at this point, so we feel very comfortable in the health and quality of the silicon. Because it goes into an automotive use case, we have to go through more rigorous testing, and all of that is underway and pretty much wrapping up, I should say, within the next month, 1.5 months. We should be pretty much done and sort of the hardware characterization testing piece of it, we have vehicles that have the silicon that are driving around. We -- the computer that actually uses the silicon is also built and well tested. We actually demoed that at AI Day last year. So we've had that in-house for over a year, too.

And more recently, it's not enough to have hardware. You have to have software running on it. So more recently, we have proven that the actual application, the autonomy application can be mapped to it. It is -- has been mapped and is actually executing as expected. We're exercising the features that are released to the public, Universal Hands-Free, Lane Change on Command, Highway Assist, all of the features, Autonomy features have been exercised and are working as expected. And not only are they working functionally, but we're meeting the performance targets we have set for ourselves. We had set a goal to at least be 4x as performance, and we are achieving that as yet without fully optimized software and our compiler tools, all of the stuff that needs to work in order to make the silicon work is all on track.

Vidya, what should investors think about for when the Gen 3 hardware architecture can make its way onto the R1 platform?

We haven't actually announced that yet. So that's something we're evaluating, but we don't have -- we haven't announced a time line on that yet. But the important thing, which I think is sometimes lost in people is that when we come out with point-to-point, it will actually be available on Gen 2 and Gen 3. so what the RAP1 chip, our custom silicon chip and LiDAR enable is really functionality at a later date. The eyes-off functionality, we've talked up for later. So there should be no difference at launch.

And for the Gen 3 hardware stack in terms of costs, how does that compare relative to Gen 2?

So part of the reason we designed it was for costs. So the way to look at it is Gen 3 will enable a few different performance points. So the lower performance point will be cheaper than the Gen 2 hardware, and the higher performance point will be on par or slightly higher, so enables -- gives us 2 different performance points and a lot of cost savings. But it's cheaper dollar per top if you look at it or dollar per performance unit.

I wanted to talk on AVs. And as a part of the Uber robotaxi announcement, the company plans to deploy AVs starting in 2028 with initial deployments in San Francisco and Miami. And I think the company is planning to deploy some R2 vehicles with safety drivers in San Francisco and Miami later this year. Maybe this one is for you, James. Can you share more on how this effort is going and the broader development there?

Yes. So I think the progress towards that later milestone is tracking very well. We had actually Uber review a few weeks ago, so I think everyone is very pleased with the progress. As you said, part of that milestone involves testing in 3 urban areas by the end of the year. So that's San Francisco, Miami and Chicago. We already have expert driver teams already driving around actually collecting data and validating the system that we wanted to show. So yes, everything looks good from that point.

And Chip, maybe you can speak a bit on the financial implications of the Uber deal, both in terms of that business on a stand-alone basis but also what it might mean in terms of selling R2 vehicles as well as the capital considerations.

Yes. So a lot of benefits in terms of the partnership. First is, of course, the capital that we're getting from Uber as a partner, where in total across the next several years, we expect to get $1.25 billion of equity capital from Uber. So we already received earlier this year an initial $300 million of capital from them as part of the signing. To James' point, later this year, we expect to achieve the second milestone, which will then unlock another $250 million, and then we'll have 3 more milestones thereafter to unlock the final $700 million. And those milestones are going to be around achieving certain technical hurdles associated with an L4-capable R2 for the robotaxi service and then the expansion of that service to up to 25 markets around the world, including at least 1 in Europe.

So from a capital perspective, one, that helps us enable the acceleration of spend on our R&D around autonomy, and that's a big part of what we need to be doing in the near term. Then over the medium to longer term, once we have an L4-capable R2, of course, we've got an initial agreement for them to purchase 10,000 vehicles, either Uber or their fleet partners, to then operate on the Uber network within another option for another 40,000 thereafter.

And an important part that we will get with that once these vehicles are out on the Uber network is the software licensing fees associated with the cars actually driving around. And I think that's an important part that many investors miss, is that a really attractive part of the economics is the fact that it's going to be Rivian's driver that's going to be driving around over time.

I don't know that you have, but any quantification of what that licensing fee would look like?

We haven't described what that licensing fee is going to look like as of yet.

Okay. Maybe talk about what the technology looks like for an L4 vehicle relative to what Rivian is currently working on for consumer vehicles, both from a hardware and software perspective and maybe how similar is it versus maybe some differences either on the hardware or software side.

So ultimately, we think the software stack should essentially be the same. So you can imagine the same stack and data loop that's ultimately powering point-to-point getting better and better and better. And as you scale up in terms of the number of data, the amount of driving actually the customer fleet is doing and you're also validating further and further over more miles, then that system gets better and better, and then you can transition it to increasingly driverless domains.

Now supporting that, you need a robust safety architecture, so that's something that very much we're thinking through. It actually comes with the RAP silicon, and there will also be some additional sensor changes for the final L4 products that we're working through the details of now. But essentially, you can think of the stack, the same stack powering the consumer fleet ultimately also powering the robotaxi fleet.

Vidya, anything?

And I would add -- yes. I think in the hardware space, what we will be doing is really having more sensing. And so I would say more and more capable is the way to look at it in order to make sure that all the corner cases are fully covered. In particular, robotaxi has to start from a stationary spot with nobody in the vehicle, which is not typically a problem you have to solve for a consumer vehicle. So that means there's a higher need for sensing around.

And so we -- there'll be additional sensing to augment the cases that are not covered in a normal consumer vehicle. And then we're just going to provision for, obviously, more compute because you want to really provision for really having all the compute you might need as the models get more sophisticated.

And I think James made an important point here for investors that the future end state is not just purely about commercial applications like robotaxi. It really is around this concept of personal L4 over time. And so the idea that all of us could collectively get our time back so that your commute now, you can be productive in ways that is not focusing on the road, like that -- for a lot of consumers, there's a lot of value to that over time.

Yes, we see -- actually, robotaxi is not the end state for us. We really think personal Level 4 is the products that we want to get to and actually robotaxi is a step along the way towards that product. So really a vehicle that you own, you can direct it to go where you want. It can go and pick things up, drop things off and really kind of work for you rather than being stuck to you.

Very interesting.

And then across the board, you would think about the cases where there's nobody in the vehicle to handle certain cases, and so there's just a higher bar on some of those things. And so we look at the entire hardware architecture and look at, so to say, what if XYZ and there's nobody, what do you need to do. So I would say everything in general is made more robust to deal with that scenario.

Okay. Sounds like you guys have a fair amount of work on all these different efforts. Another one I wanted to touch on briefly is robotics, another physical AI application. Rivian spun off Mind Robotics. How much of what Rivian is working on is applicable to the field of robotics? Is that something that your teams are engaged in?

So I think the jury is still out a little bit on how much these models transfer directly to the physical AI space or not. I would say that the idea of using human-derived ground truth is very powerful, so the imitation learning, and that's something that our models are trained to do. Now in the robotics space there's, I mean, 2 additional problems. So one is the actuation is much more complex, so there's a lot more degrees of freedom. And it's actually not clear how to get the human-derived ground truth from that.

And then it's also -- you have some more ambiguity in the classes and the types of problems you have to solve. So there's more generalization problems you have to solve versus driving, which is really the one task. So I think the teams talk closely together on what are the best techniques. I don't know that they'll be the same right now.

And I would say on the hardware side, the chip we built is really the first application was autonomy, but it's a very capable chip for physical AI. So if anybody out there is interested in using our chip for robots, do reach out. But we -- definitely is an area of potential engagement. Of course, we can't speak for Mind. They're another company, but...

Okay. That makes a lot of sense. I want to talk about the collaboration with VW and you have the joint venture with them. Maybe talk about how the integration of Rivian's electrical architecture is going with respect to VW and time line for when the first VW vehicles may launch using Rivian technology.

Yes. So in terms of the -- and VW has discussed this externally. They expect to roll out their first vehicle that includes the electrical architecture and software next year, which will actually unlock the last equity investment by VW. Overall, the partnership continues to be going very well and looking forward to having their first vehicle out on the road.

And what is the company's current thought around potentially licensing the EE architecture to other auto OEMs?

Yes. Look, I think always open to have conversations. And I think RJ being a very prominent member of the automotive leadership community is talking with any number of other OEMs. And I think part of what the industry has learned over time is that this shift to a software-defined vehicle is where consumers want things to head over time.

And sort of the legacy architecture of outsourcing every single black box in the software to Tier 1 or even a Tier 2 provider, as you think about that software-defined vehicle that we're developing now and have now, there are limitations with that, that you can't offer with what we can now do. And so certainly, it's a latent opportunity, but our biggest focus right now is continuing to execute both in terms of our road map with R2 but then also executing with our road map with our partners at Volkswagen.

I do want to spend a couple of minutes around battery technology. I think it's a really important part of the long-term opportunity with electric vehicles and some really fascinating stuff that's underway. For R2 using 4695 cells, maybe talk about why Rivian made the choice to use 4695.

Yes. So it's -- one, it's about higher power density, but then also from a structure of the R2, we now have a structural battery pack. So you've got fewer battery cells within the battery as well. And so a lot of it, as we think about driving over the long haul, it's about how you drive down costs for manufacturing R2.

And any update on when Rivian will source those from the U.S.?

So the plan is still with LG to start sourcing that next year.

There's also innovation beyond the cell form factors in terms of ultra-fast charging cells, and there have been some products that have been announced recently where batteries can charge in 3 to 5 minutes, which I think would be kind of game changing for the EV market, including in the U.S., for people who like to road trips. I'm curious what Rivian's thoughts are about potentially using that kind of ultra-fast charging battery technology over the medium to longer term.

Well, look -- and certainly, either of you jump in here as well. I think the question is the use case for customers and what's the necessity there, and we were talking about this earlier today in some of our conversations. But for the vast majority of owners of EVs in the U.S. are plugging their cars in overnight and charging things up overnight.

Now certainly, there are use cases where you're doing the long-distance drive going across the country, but current technology can cover the vast majority of those use cases. And even I will say personally, when I'm driving with my family in our EV going on that longer road trip, it's actually nice to be able to stop and get out for 20 minutes and let the kids run around, go get a Starbucks, go get a sandwich.

So I think part of what a lot of consumers, until you've owned an electric vehicle, don't realize is it's just a different behavior pattern for how you're going to treat a road trip. It's no longer, okay, we got to the fill up the gas tank as quickly as we can and get back on the road. It might be slightly more leisurely in that you're waiting to charge for 15 or 20 minutes. But at least in the case of my family, I actually think it's a very nice and different way of approaching a long-term drive.

Vidya, I mean, just from an engineering perspective on the hardware side and what would have to happen to integrate those types of cells, I mean, how easy or hard might that be?

I think -- I mean, I look at the analogies like LiDAR. LiDAR technology existed for a while. The question is always when are you integrating it. And for the longest time, it was too big and too expensive, right? And I think when we just integrated now for R2, it was sort of both of them were right. The price was right and the form factor was right. And I think the same with alternate technologies when the sort of the price is right and the use case is right sort of justifies it. I think we can definitely look at it.

And I think we've had a lot of experience dealing with battery management systems. That's where all the complexities moved to the BMS. And I don't think that's easily a solvable problem. I think it's the availability of the technology at the right price.

Very interesting. A handful of financial questions and maybe, Chip, I'll direct these at you. The company is expecting a 50% reduction in bill of materials for R2 relative to R1. Maybe talk about to what extent that's on track, especially with all the inflation around materials, semiconductors and other components.

Yes. I mean that goal to get there is as we've fully ramped up R2 at our plant in Normal. So still relatively early days in terms of that ramp, as we talked about on our most recent earnings call. In the second quarter, we did incur about $100 million of incremental expense through COGS associated with the ramp of R2. And we, as we said on the call, expect to continue to see more ramp costs flowing through here in the third quarter before in the fourth quarter then flipping the other way and being on our way towards achieving automotive gross profit positive on an exit rate basis for the year. So that still remains the goal, but we're still early in our ramp.

And as you just alluded to, the company does have that positive automotive gross profit target exiting this year. What are the key things that need to happen to get there?

Well, it's -- we haven't given a specific volume number, but it really is around continuing to see our supply chain mature, continuing to have them ramp so then we can ramp within the plant in Normal because a big part of what is going to help our automotive gross profitability is both delivering more R2s but then also seeing that fixed cost leverage within the plant as well, where as we start then sharing the shared costs across our paint shop and our stamping press across a much broader set of manufacturing across R1, R2 and EDV, that helps the profitability not just of R2 but also the other vehicle lines.

So scale kind of being one of those main drivers to getting there, you guys aren't making any unusual assumptions around pricing or cost reductions or anything like that?

No, correct. Yes. All about scaling and building more R2s.

Well, maybe bridge us from sort of that target of positive auto gross margin exiting this year to the medium- to longer-term positive EBITDA. And I know the formal target there net with some of the extra spending you're making on autonomy has been removed. But what are some of the key parameters investors should think about as Rivian ultimately works toward that positive EBITDA number?

Yes. Look, I think it's going to be driven by several factors. One is our ramp of R2 and how successful and how fast we are in the ramp of R2 plus then the demand for R2. Second then is our continued investments in R&D and SG&A over time and what are those going to look like. So we haven't provided a new long-term target, but it's going to be based off of those key drivers.

Okay. On the balance sheet at the end of 2Q, the company had $14 billion of cash and available liquidity. Talk a bit about how the company is thinking about managing the balance sheet.

So that was -- it's not cash and available liquidity. That's cash and then the investments that we expect to receive. Plus then our Department of Energy loan gets us up to the $14 billion. But we do have that line of sight for funding the business. And so in terms of managing the balance sheet on a go-forward basis, first and foremost, it's going to be continuing to execute on the ramp of R2 because that's then going to drive what our profitability ends up looking like and the funding needs associated with that.

Second then and part of the reason why we raised capital back in July is around funding the equity commitments and then any reserves around the plant in Georgia. So we feel as though we'll continue to remain opportunistic should we need to raise additional capital but have a good line of sight for a significant amount of capital for the next good period of time here.

A couple other questions I wanted to ask related to R2. The company's talked about selling that in Europe eventually. Can you share more on your current plans, when that might happen? And does the penetration of the Chinese companies into Europe alter that calculus at all?

Well, look, we know R2 and even R3 eventually will be very well priced for the European market and are sort of -- given the size and form factor for what they're looking for in Europe could be a very good opportunity. Our focus right now remains on continuing to ramp here in North America, first and foremost, recognizing that it is a competitive landscape, and there are some very interesting vehicles coming into the European market. But we want to make sure we're executing here first.

And as you think about the software side, potentially selling into other geographies like Europe, obviously different regulations, maybe different driving patterns. What would you have to do if you're going to offer some of these autonomy features in Europe from your team? How similar or different might it be?

Yes. So the current approach is we take maybe 10:1 customer data to expert driving data. And then the reason we need that expert data is to kind of dial in better behavior. So customers are often not stopping at stop signs fully. They may be speeding slightly in some cases. So we actually want that expert data to kind of fine tune, hone in a good, safe Rivian driving experience.

So we would deploy in a similar fashion that expert team in Europe wherever the new country is, collects an amount of data there. And when we've done that in the past, we've actually seen good transferability. So then we add that to the pool and then you fine tune. So that would be essentially the approach. And we think that with relatively modest coverage of new road conditions and new road rules that we can actually adapt the system fairly quickly to those new cases.

One other thing I wanted to close on was around autonomy for commercial vehicles and just some of the broader things maybe your team is working on there. I know the EDV business has been growing pretty nicely. But to what extent are you working on new products for that use case?

Yes. So I think in general, again, we see that stack as being very general. We've built flexibility into the system kind of from the get-go. So we can move sensors around. We can add and remove different modalities as needed. And actually, the system largely stays the same. So now we would, in a similar way, need to get a certain amount of fine-tuning data for new form factors. But we think ultimately the driving task is pretty similar in all these cases and the system should generalize well.

And have you guys given any time frame for autonomy products in the commercial space?

No, we haven't. But it's something we're always discussing internally.

Great. Well, I learned a lot. Thank you all for taking the time to speak with us. Appreciate it.

Thanks, Mark.

Thank you.

Thank you.

Good afternoon, and thank you for joining us for Rivian's Second Quarter 2026 Earnings Call. Today, I'm joined by RJ Scaringe, our CEO and Founder; Claire McDonough, our Chief Financial Officer; and Javier Varela, our Chief Operations Officer.

Before we begin, matters discussed on this call, including comments and responses to questions, reflects management's views as of today. We will also be making statements related to our business operations and financial performance that may be considered forward-looking statements under federal securities law. Such statements involve risks and uncertainties that could cause actual results to differ materially. These risks and uncertainties are described in our SEC filings and the earnings presentation we filed with the SEC today.

During this call, we will discuss both GAAP and non-GAAP financial measures. A reconciliation of historical non-GAAP to GAAP financial measures is provided in our earnings presentation and press release.

Just before the earnings call, we posted our earnings presentation, which includes an overview of our progress over the recent months. I encourage you to read it for additional details around some of the items we will cover on today's call. Following our prepared remarks, we will be taking questions from sell-side analysts. In the interest of keeping the call to 1 hour, we would ask these analysts to limit any follow-on questions to one.

With that, I'll turn the call over to RJ.

Thanks, Chip, and good afternoon, everyone. Thanks for joining us for today's call. In the second quarter, we began deliveries of our 2 vehicles to external customers. We expect R2 to be a game changer for our customers and a driver of Rivian's long-term growth and profitability.

Importantly, in the U.S. automotive marketplace start for high-quality EV choice, I strongly believe R2 is an attractively priced option for everyday ventures that will resonate with a broad set of consumers. Media reviews of R2 are outstanding. One journal wrote, "R2 is one of the best new cars I've driven in years." Another praise R2 as among the finest vehicle designs in the world. Early customer feedback of R2 is positive, and I have to say, I love it as my daily driver.

Strong R2 reviews and increasing brand awareness are directly translating into accelerating customer touch points with our products. In the second quarter, we hosted over 57,000 demo drives, a Rivian record. While it's early in our ramp, we are pleased with the reservation conversion we've seen to date for launch addition R2.

Turning to our manufacturing footprint. Our team in normal is focused on the production ramp of R2. As a reminder, R2 production started with a single shift, and we expect to scale to 2 shifts by the end of the third quarter. We are seeing good progress in new team member training and process improvements for the ramp of the first R2 shift.

Additionally, members of our supply chain team are spending time on site with suppliers as we look towards supporting the ramp in the back half of the year and into next year.

In Georgia, construction of our next manufacturing site is progressing well. The combination of our normal and Georgia facilities provides Rivian the path to scale to 515,000 units of annual capacity with plenty of room for future expansion.

As an update on our Amazon partnership, Amazon now has over 40,000 Rivian electric delivery vans active in its fleet, delivering packages across thousands of cities in North America. In addition to the EDV standard pack variant already on the road, we're developing new variants with a larger battery pack and all-wheel drive to support Amazon's needs. In the second quarter, we surpassed 1 billion miles driven on the Rivian commercial van platform.

Turning to our technology road map, Autonomy remains one of the most important areas of investment at Rivian. In the not-too-distant future, we believe advanced assisted driving capabilities will be a key differentiator for our customers and a substantial driver of market share. Our autonomy development is on track, and we expect to begin rolling out point-to-point capabilities by the end of this year.

We believe our rapidly expanding car park with R2 will accelerate our data flywheel to enable delivering hands-off, eyes off capabilities in 2027 and ultimately, Level 4 capabilities in consumer and robotaxi variants of R2.

Since we began monetizing Autonomy us in April, take rates are trending positively for the service, and we believe further future releases will drive continued increased uptake of Autonomy Plus. Later this year, we will host another autonomy and AI Day where we'll provide more updates around our exciting autonomy and AI progress.

At the end of 2026, we are targeting the launch of our third-generation autonomy hardware -- this includes RAP1, our first in-house design silicon and the addition of LiDAR. The development of our RAP1 chip is on track, and we've begun final testing phases of production silicon. Finally, in May, we rolled out our Rivian assistant on all R1 vehicles and plan to launch the feature on R2 with an over-the-year update later this year. Rivian Assistant is our new AI-powered voice assistant that is built to be a digital co-pilot with integration into the vehicle ecosystem and other external apps. Customers love this new future and its intuitive functionality across applications like Spotify and Google.

In closing, I want to thank our team for their continued execution this quarter. We are laying the foundations for a future that we believe will be fully electric, autonomous and AI defined. The reviews for R2 affirm that our category-defining brand and vertically integrated and extensible technology resonates deeply with customers. I couldn't be more excited about the upturn ahead.

With that, I'll pass the call over to Claire to discuss our financial results.

Thank you, RJ, and good afternoon, everyone. As RJ shared, we started external customer deliveries of R2 in June. My family took a road trip in R2 earlier this summer, and we put the storage capacity to the test by piling in 4 large suitcases, backpacks, golf clubs, tennis rackets and groceries. Road tripping with universal hands-free is outstanding, making long car rides relaxing. I'm still looking forward to the launch of point-to-point capabilities later this year.

As I discussed previously, delivering a strong 2026 exit rate for R2 production and deliveries is a key focus for our team as we believe it will directly translate into positive automotive gross profit for the business.

Turning to the results for the second quarter. Our consolidated revenue was $1.66 billion, a 27% increase over the same quarter last year. Consolidated gross profit was $179 million, and our gross margin was 11%. Gross profit included $138 million of depreciation and amortization and $31 million of stock-based compensation expense. Adjusted EBITDA losses for the quarter were $379 million, driven by our $179 million of gross profit and increased adjusted operating expenses as we prepare to scale R2 and invest in our autonomy road map.

In the second quarter, we produced 12,613 vehicles at our plant in Normal, Illinois, and delivered 12,194 vehicles. Delivery results topped our 9,000 to 11,000 vehicle expectations due to robust growth quarter-over-quarter in EDV and R1, coupled with the introduction of R2 deliveries. Our $1.14 billion of automotive revenue, a 23% increase over the same quarter last year were primarily driven by a 14% increase in vehicle deliveries and $103 million increase in revenues related to automotive regulatory credits, partially offset by a decline in average selling prices from a higher mix of commercial land and R2 deliveries.

Automotive gross profit loss was $36 million compared to a gross profit loss of $335 million for the same quarter last year, a $299 million improvement, primarily due to increases in delivery and production volumes and increase in revenues related to automotive regulatory credits and an IEEPA tariff refund receivable, partially offset by the ramp of R2 production. In the quarter, we recognized approximately $100 million in incremental cost of revenue due to the ramp of R2 production as compared to the production at more normalized levels.

While current macro and geopolitical factors are creating added complexity, cost and uncertainty, our team continues to work hard to manage supply chain risks and increasing commodity and memory costs.

Our Software & Services segment reported another strong quarter. During the second quarter, the segment generated $515 million of revenue, a 37% year-over-year increase, primarily due to an increase in vehicle electrical architecture and software development services by the joint venture, repairs and maintenance and Autonomy Plus, offset by lower remarketing sales. $8 million or 60% of software and services revenue was attributable to our joint venture with Volkswagen Group. Software & Services gross profit was $215 million, a 42% margin due to higher vehicle electrical architecture and software development services by the joint venture.

Looking at our balance sheet, we ended the quarter with approximately $5.3 billion of cash, cash equivalents and short-term investments. In July, we sold 86.25 million Class A shares in a follow-on equity offering to raise approximately $1.3 billion for general corporate purposes, including the funding of equity commitments and reserves for our $4.5 billion Department of Energy loan associated with the construction of our Georgia plant.

Additionally, later this year, we expect to receive $1 billion in nonrecourse debt from Volkswagen Group and an additional $250 million equity investment from Uber, both subject to the completion of certain conditions. Our total available liquidity and targeted future capital is over $14 billion to support Rivian's investments in growth initiatives. This includes current liquidity, the Department of Energy loan and additional targeted equity investments from Uber and Volkswagen Group, which are all subject to certain conditions.

Finally, for our 2026 guidance, in early July, we announced we expect to deliver 65,000 to 70,000 total vehicles across R1, R2 and our commercial van, a 3,000 unit increase as compared to our prior guidance. This implies approximately 42,400 to 47,400 deliveries in the second half of the year. We expect deliveries will be weighted towards the fourth quarter as we ramp R2.

While we continue to believe our gross profit will increase year-over-year, we expect the complexity of a new vehicle launch will negatively impact our automotive gross profit in the third quarter as it did in the second quarter before becoming a benefit for our overall operations in the fourth quarter as we ramp production and deliveries. As a reminder, we believe this is a transition year for the automotive segments path towards long-term profitability as we scale R2.

For 2026, we expect an adjusted EBITDA loss of between $2 billion to $1.8 billion, a $50 million improvement at the midpoint due to better-than-expected revenue related to regulatory credits in the second quarter and increased delivery volumes, which were partially offset by raw material, memory and logistic cost increases.

While economic and geopolitical conditions, including international conflicts pose risks, we remain steadfast in our plans to invest behind key growth drivers. We continue to progress our autonomy road map and the expansion of our sales and service footprint as we scale with R2. We believe these strategic investments will deliver long-term value to our shareholders and customers.

Finally, for 2026, we are reducing our capital expenditure guidance by $250 million at the midpoint to $1.7 billion to $1.8 billion, benefiting from project efficiencies and timing of spend. Our CapEx spend is primarily related to finalizing construction and tooling for R2 in Normal, the continued build-out of our sales, service and charging infrastructure and construction for our greenfield plant in Georgia.

In closing, I'd like to congratulate our team again for the start of external R2 sales in June and the continued strong execution in the second quarter. We continue to believe that R2 and our technology road map will be truly transformative for the growth and profitability of our business.

I'd like to turn the call back over to the operator to open the line for Q&A.

[Operator Instructions] Our first question comes from Mark Delaney with Goldman Sachs.

Congratulations on the R2 launch. Very exciting to see. I was hoping to start there on R2, can you share more on how the production ramp is progressing? And any examples or metrics maybe you can share on that front to illustrate how the ramp is going? And then as you think about R2 over the next year, think the pace of deliveries is going to be gated more by a supply or demand?

Well, thanks, Mark, for the question. We're also very excited about R2. Well, I guess, first on the production ramp. Of course, ramping production of a vehicle, especially where it's a first-time launch of a new vehicle is complex. It's an orchestration of ramping hundreds of suppliers, the suppliers all need to ramp at the same rate. Of course, our overall production output is throttled by our gated by the slowest moving supplier. And so the coordination to make sure everything is ramping consistently and in an organized fashion across all the different suppliers is really key focus for us. And it's the reason we start on a single shift and get everything moving efficiently and effectively before bringing on a second shift.

So saying that, this is a complex effort. There's lots of moving parts, and we went into it fully recognizing the complexity and having been through this with R1 and iterations in R1 with a lot of learnings and knowledge of a business that we certainly didn't have to this level with the launch of R1. And so where it sense today is we're -- we are ramping. We're continuing to make progress week over week. We've -- as you saw, we've updated our guidance to reflect the progress we're making. And the biggest category of focus for us has been supply base and making sure that all of our suppliers are ready to continue ramping to much higher levels as we look at the back half of the year. And in particular, as we start to plan for running -- fully utilizing the plant and running across 2 shifts.

Now on the demand side, there's an enormous excitement around the product. And we've been really encouraged by the conversion of reservations to orders on our launch addition. We, of course, had our own internal projections as to what that would be, recognizing that we're launching with the launch addition, which is a $58,000 version of the vehicle, and most expensive version of R2. And that conversion rate from reservation to orders has been meaningfully higher than what we expected, which is a great signal. Now of course, with that said, we are looking forward to introducing the other trims our mid-spec and our base or standard spec as we look at the early part of 2027. But the excitement even for the launch addition and the higher price finance addition has been very, very high.

Appreciate all those details. My other question was trying to understand what the R2 ramp might mean for the financials and putting that into context with some of the other comments around the input cost environment. So as you look out through the end of the year and start ramping R2, does Rivian still expect to reach a positive gross margin exiting 2026?

Thanks, Mark. As I mentioned in my prepared remarks, we continue to expect that R2 will achieve a positive gross profit as part of our exit rate for 2026. And that's enabled, as RJ mentioned, by the ongoing efficiencies of being able to produce R2 at much higher volumes and the fixed cost leverage that, that will enable within the business.

Your next question will come from George Gianarikas with Canaccord Genuity.

I wonder to focus a little bit on the EV. Can you provide an update maybe on commercial adoption beyond Amazon? And as you expand pilot programs for non-Amazon fleets, what's the current primary governor on deployment pace? Is it TCO parity, upfront capital outlay, charging density, anything you could share there?

Thanks, George. I think the first point to just call out is the ramp-up that we're seeing within Amazon. And so you're seeing this already in the numbers for the first 2 quarters of this year that Amazon is purchasing more vans. We're ramping up meaningfully from where we were in previous years. And that's absolutely a reflection of the TCO advantages that our EDV platform represents and enables. And we think that those cost benefits that are being so clearly realized by Amazon and so clearly evidenced by the ramping up of volume, we think that's going to translate to other commercial operators ultimately seeing those cost benefits and recognizing both to remain competitive, but also to address their own cost structures that they're going to be pursuing opportunities with EDV.

Now that has taken longer than we had would have expected or we would have hoped for in terms of other fleets adopting a strategy towards electrification and a strategy of adopting not just electric vehicle, but a highly connected intelligent vehicle. But the signal that's being sent through Amazon's significant ramping of this is really outstanding in terms of a billboard for the capability and a billboard for what the EDV is possible delivering from a business point of view.

And maybe more of a long-term question. As you continue to benchmark other global EV platforms, particularly from Chinese OEMs, what structural or operational lessons can you apply to the R3 and other future vehicles as you spec those out and try to build them?

Of course, the auto industry is a business where every manufacturer is very aware of what our other manufacturers are doing. So Rivian are purchased by probably every car comp of the world and taken apart. And similarly, the Chinese vehicles are purchased by many different -- both OEMs, but also benchmarking companies. And so as a result, there's not really -- there's not like a magic difference in terms of how a Chinese vehicle is built relative to a vehicle built in the West in terms of the manufacturing approaches, the joining technologies, the use of high-pressure die castings, consolidation of stampings. Those methods are being deployed across all, call it, best-in-class vehicles from a manufacturability point of view and a vehicle architecture point of view. We see that, of course, evidenced in how we're building the R2, use of high-pressure die casting is an extreme focus on part consolidation and part elimination.

I think the big difference that exists between U.S. cost structure and Chinese cost structure is the input cost. So you have a much, much lower labor cost in China. Much, much, much lower capital cost structure where, in many cases, the capital cost is zero, meaning it's being provided by the local government. And that compounds across the supply chain and all the way up through the OEM, and it results in a materially lower cost than what it would take to produce in the Western world. And so that naturally opens the question around trade policy. And importantly, it opens very complex questions around supply chain strategy. And we're navigating a lot of those complexities around sourcing our components and our -- not just our Tier 1, but even down to the raw materials, where there are coming from recognizing the frameworks that have been established around trade.

And in a world of completely open trade, it would look -- our supply chains would look very different because we'd be optimizing purely around cost, which would be, of course, optimizing around countries that have the lowest input cost structure, be it labor cost, lowest energy cost, lowest land cost, lowest cost of capital. And of course, that's not the world we work in. We work in a world where there's very intentional industrial policies that exist that changed the way we might be able to look at this. And in many cases, there's implications around certain components where we believe it's very helpful to make sure those are sourced from the United States.

Your next question comes from Shreyas Patil with Wolfe Research.

I was curious if you could maybe unpack the $50 million increase in adjusted EBITDA for the year. How much of that is tied to the better regulatory credit sales? And then how should we think about input cost inflation?

Sure, Shreyas. As you look at the evolution of our EBITDA guidance, so we were able to improve the bottom end of the range by $100 million that moves the midpoint $50 million, as you stated. We had higher-than-anticipated regulatory credits in the first half of this year. And then as we look ahead, we have the added benefit of the incremental delivery volumes with the increase of 3,000 units in our outlook. But offsetting that, we have headwinds associated with increases in raw material costs, increases in memory costs as well as increases in logistics costs that had us arrive at this specific EBITDA range and outcome.

And so as we look at the second half performance relative to the first half, what you can anticipate seeing is on the gross profit side, as I mentioned in my prepared remarks, we'll continue to see some of the inefficiencies and complexities associated with the R2 ramp impact Q3 automotive gross profit, and then we'll start to see some of the tailwinds of the scale and volumes kick in, in the fourth quarter. And then beyond that, we'll see just the ongoing investments that we're continuing to make across R&D and specifically the growth in our autonomy spend and then the continued fixed cost leverage that we'll see as we're growing our sales and service footprint, but certainly growing sales much more rapidly as well.

Okay. And then maybe just on the Autonomy Plus. So you're going to be launching the point-to-point solution later this year. I'm curious where things stand today with Autonomy Plus. Just if there's a way you can give us a sense of take rates or kind of adoption trends. And then how do you see point-to-point driving both demand, but then also those take rates? So ultimately, the subscription revenue opportunity.

We've been really encouraged by the take rates on Autonomy Plus to date. And I think, in particular, what's really encouraging is to see the status that we're at there in terms of customer engagement, customer usage of the future and then, of course, related to that, the take rate around paying for Autonomy Plus. But we have very clear line of sight and visibility into what's coming. And so as encouraged as we are with the take rates today, we're quite confident that, that will expand meaningfully as we add in additional capabilities and features.

And there's going to be a whole -- think of it as like a staircase of features that ultimately leads us to very high levels of autonomy, but a really important step that's happening in the very near future is the capability of point-to-point. And just to restate what that entails, that means you can get into the car, type the address into the vehicle and the vehicle completely drive itself to that address. And initially, that will be a feature in which it's like Level 2++ where your hands are off the wheel, but your eyes are still primarily on the road, meaning you can look away for short extended moments of time, but you're not in a position to officially be reading a book or be responding to e-mails.

Now this step in 2027 is going further where we go, not only hands off, but also eyes off, and that's really important because then not only is the vehicle fully driving itself and navigating to the address, but importantly, you're able to truly get your time back without the vehicle dinging it, you say, to keep your eyes on the road, but actually be able to read a book, be on an iPad. And so that's really into the like true as per the definition, Level 3 capability.

And then, of course, into 2028, that's when we bring on Level 4. And that's where the vehicles capable of operating as a robotaxi, meaning there's no one even needed to be in the driver seat for any operation of the vehicle, and you can put occupants in the back. You can use that capability for robotaxi applications. But you can also have that, which is one of the things we've been very excited about, have that as a really nice feature on an owned vehicle. And so we think of this as personal Level 4, where your own personal vehicle can drop you at the airport, it can pick your kids up from school. It can go to the grocery store and grab things for you. And that road map is something we're not only very excited about, but we think will fundamentally drive increasing levels of adoption on Autonomy Plus and its future variances we've just described.

Now with all that said, we also think we're at an inflection point from a consumer behavior and consumer mindset, where historically, autonomous features were not the primary purchase criteria for most customers. Of course, there's technology early adopters for which this has been top of mind for them for some time. But for the vast majority of customers, this wasn't the driving purchase criteria. We think this starts to emerge to be much more of a primary purchase criteria. And it's one of the reasons we've made the decision to invest so heavily here.

We think ultimately, market share is going to be driven by the capability of the vehicle to drive itself. And of course, linked to that the vehicle being software and AI defined. And those are 2 very like critical pillars of our technology road map and technology strategy.

Our next question will come from Rajat Gupta with JPMorgan.

I had a question on the R2 customer profile. Any early reads on what that customer is proving to be existing electric vehicle owners, first-time electric wafer buyers, best lacrosse shoppers? Just curious if you have a read on the customer profile of the early R2 orders, and I have a quick follow-up.

Thanks, Rajat. The goal in developing and designing R2 is to have a very diverse set of customer profiles, meaning customers that were coming out of a wide variety of form factors coming out of a wide range of brands and from a broad spectrum of demographics. And that was the intent, that was the target. We now have the benefit of being able to see the data to see who is actually buying R2s. And the great news here is that it is exactly that. We have a very diverse customer set coming out of a broad range of vehicles, which I guess perhaps isn't -- shouldn't be too big or a surprise, the midsize SUV segment, the 2x5 passenger segment is the largest segment by far in the U.S. market. It's generally the largest segment globally as well.

But the way that we've been able to really thoughtfully execute this vehicle in terms of packaging, storage capacity, performance, range, efficiency driving dynamics, both on-road and off-road has created a very wide net for attracting this very diverse set of buyers. And maybe one of the most exciting metrics here is we have a significant number of first-time EV owners, and that's really powerful. So for someone to have their first time owning an EV be Rivian. And for that bond with the brand, so to speak, to be created through that first-time ownership is powerful. But it also illustrates the strength of what we've created in terms of the product and the product market fit that this is mobilizing demand for EVs that's been latent, that's been sitting on the sideline because the product that sort of would motivate someone to move from an ICE to EV hasn't existed until R2. And so this is a great metric.

Now of course, there's -- there are EV buyers that are in the mix. You have folks coming in a Tesla or folks going out of Rivian R1s, but the most important metric here is the very significant number of folks that have not owned EVs before.

That's great color. And just a follow-up for Claire, maybe. I noticed in the second quarter gross profit commentary, you mentioned on IEEPA tariff refund benefit. I'm curious how much that was because it didn't see it in the full year outlook. And then just relatedly, it looks like the guidance implies a steeper EBITDA loss in the second half versus first half despite the positive gross margin exit rate commentary. Just trying to bridge those 2 aspects.

Sure. On the IEEPA tariff refund piece, as we mentioned in our earnings call last quarter, we size the IEEPA tariffs in the tens of millions of dollars. The receivable that we booked this quarter is roughly half of the total. And so the receivable is related to the refund of the amount of anticipated proceeds that are not subject to the existing DOJ appeal.

And then as you think about the second question on the EBITDA loss being steeper in the second half of the year. One of the big differences there is we did have $164 million of regulatory credits that benefited our gross profit outlook in the first half. And so as we look at the second half of the year, we'll see, again, some of the complexity associated with the R2 ramp impacting automotive gross profit in the third quarter before showing progress in the fourth quarter. But the bigger driver is really the absence of regulatory credits in second half versus first half and then the ongoing investments in R&D and SG&A.

Your next question comes from Itay Michaeli from TD Cowen.

Great. Just first, a couple of questions on the R2. RJ, 57,000 demo drives. It's a big jump sequentially Hoping you could talk a little bit about what you're seeing in terms of conversion of those drives into orders? How many people are actually experiencing autonomy plus? And then secondly, just curious whether you do expect to deliver a material amount of the premium trends this year or R2 sales just predominantly be the performance trend this year?

Yes. I mean we're really excited, as I said in my opening remarks, the 57,000 drives this past quarter is a record for Rivian. And so just the delivery of that, just enabling that number of drivers you said roughly doubling where we've been in the past. The team has had to really come together and we do a lot of prep knowing that the amount of excitement on R2 is going to drive a significant increase in both exposure to the brand, touchpoints of the brand, but of course, that's manifested through these test drives. And so with that, we're not providing specifics around conversion, but that's really an important metric we look at internally. And that actually links pretty heavily to the availability of the other variants.

And so to start, as you said, we have our performance brand the most expensive variant of R2 that we've launched with. We'll be introducing the premium and then our standard trims as we look at early 2027. But we are witnessing a higher level of conversion than we expected, meaningfully higher than we expected into the launch edition package, which was great. It's super encouraging. I think reflects just the amount of excitement around the product. but saying that we fully recognize and are excited to also be launching the other trims and know that those will be coming, as I said, at the early part of 2027.

That's great to hear. And then Claire, maybe just a quick follow-up. R&D, I think, was flat up sequentially. I know you're making a lot of investments in autonomy. Just curious how we could think about R&D kind of trending in the second half of the year.

Sure. As we look at the second half of the year, we do anticipate seeing a step up and that's predominantly related to sourcing GPU for our autonomy training that will continue to ratchet up over the course of the second half of the year. In the second quarter, there, we saw some level of production as we're now in production. So there was a decrease in our engineering design and development spend, but we did underneath the R&D spend still see significant increase in terms of our overall autonomy spend.

Your next question will come from Joseph Spak with UBS.

Just maybe the comment about the $100 million of incremental cost of revenues of ramp of R2 compared to normalized levels. Just maybe some context, like what are you classifying as normalized? Is that 1 shift, 2 shifts? And are you also assuming a more normalized mix in that analysis?

Joe, the way to think about it is the $100 million reflects a combination of ramp-related inefficiencies such as expedited freight costs out of the gates or short-term premiums with suppliers that we anticipate coming off as we increase our production volumes over the course of the next couple of quarters. And then it's also cost that we expensed in period that would have naturally been absorbed into inventory, for example, if we were at more of a Q4 run rate of production volume. So that's a little bit of the way to think about how we've sized the $100 million of estimated ramp costs.

Okay. So yes, that's what I -- I mean, a good portion of that is just like you don't have the scale and leverage over the assets. So if you're saying second shift comes on at some point, later in the year and you're exiting gross profit positive, then that's sort of the -- that could help us sort of triangulate sort of like the units do you think you need to sort of be close to breakeven. But that's a fair way to think about it.

That's generally a fair way to think about it. And so as we talked about it being really at sort of Q4 normalized levels, that's not the full extent of R2 fully ramped up. But is, in the near term, a more normalized production environment that we can compare our standard cost too.

Okay. And then RJ, I appreciate the Autonomy Plus commentary earlier. I was wondering if you could also give us a latest update on just RAP1 and Gen 3 hardware and any progress? Is that still on track for SOP later this year?

Yes. Our in-house silicon RAP1 is on track. As you know, that underpins what we call our Gen 3 autonomous hardware platform. And so both the RAP1, along with the rest of the upgrades to the system are on track for later this year.

Your next question will come from Dan Levy with Barclays.

Okay. Great. I just wanted to start with a question for you, Claire, on just unpacking the COGS per unit on Joe's question there. To only see COGS per unit up a few thousand dollars quarter-over-quarter despite that incremental $100 million says that there's probably something incrementally that's going on quarter-over-quarter on the cost side, whether it's within EV or one. So maybe if you can just help unpack if there's other underlying cost improvements and why we're not seeing maybe -- what should have been maybe a step-up in raw material pressure in that cost per unit?

Sure. As you look at the quarter, we had $96,700 a unit of COGS. If you were to back out the $100 million on a per unit basis, that's about $8,200 a unit that would have gotten us to $88,400. That's about a $5,000 per unit improvement on a quarter sequential basis relative to where we were One of the biggest drivers of that improvement is through the volume that we saw in the quarter as well. So that's going to be -- continue to be a key theme as we'll see our COGS per unit continue to come down over the corresponding quarters of this year. And then beyond that, we did also have at the IEEPA receivable that was also benefiting this quarter.

That would have been [indiscernible]? Or was that booked in COGS?

Booked in COGS.

Okay. As a follow-up, RJ, I want to ask something that -- I know you've addressed in the past, but I think it's important just in the early days of the R2 ramp. If we look back to R1, the days of R1 and early ramp, there were -- you probably bid off a bit more than you could to, which is something I think you've acknowledged. There's a lot going on, a lot of different configurations. And we look now, I know you talked about focusing on just the launch of tradition. But broadly within Rivian, there are a lot of different initiatives between RAP and I'm sure there's multiple other future iterations planned, your Autonomy road map, what's going on with network architecture and then also just Mind also just a number of different initiatives here. So maybe you can just talk us through how this time is a bit different versus R1, how you're approaching R2 execution given that you still have a lot of things on your plate and how you're looking at resource allocation in this environment?

I mean the -- it does seem in some ways like yesterday, we were launching R1, and then it also seems like it was a very, very long time ago. But with that said, the company that's launching R2 today is a very different business than what launched R1. I say that with regards to the maturity of our processes, the maturity of the team, the maturity of the development sequencing. And when we look at the launch and what we're going through today in terms of ramping production, ramping suppliers, what we're experiencing today is actually an artifact of what we did in the 2 or 3 years leading up to it. And so a big difference between R1 and R2 is the structured approach we took to what we call design validation builds. So these are vehicles that were in production spec, but built on our pilot line. We went through multiple iterations of that.

And then subsequent to that, our manufacturing validation builds where it was, of course, the production spec, but built in the manufacturing plant. And the MVP builds, those manufacturing valuation builds started at the very beginning of this year, at the beginning of 2026. And so that gave us time to work through issues around ramp, work through issues around design elements, of course, supplier-related issues as we're doing those validation builds. And so the number of unexpected surprises in now the production ramp-up of R2 is far lower incredibly -- it's an incredibly different situation than what we had in R1.

Now that's not to say there aren't surprises. It's -- as I said at the start, building and launching in new cars, just an inherently very, very complex activity. And it's complex if you're a company that's been building cars for 100 years, it's complex if you're Rivian. And so we've done everything we can to prepare. But as we now are going through the ramp, we're still being very thoughtful around how rapidly we ramp up our supply chain, making sure that all of our suppliers can ramp consistently, planning that very thoughtfully before we bring on the second shift. And a big enabler to help make that ramp process more seamless and more straightforward was the simplification of the product relative certainly to what we did before.

As you said in R1, we didn't just launch our one. We launched R1T, R1S, and a commercial van, all within the same 3-month window. And then we had a very complex set of build variations essentially thousands of different possible build combinations for the R1T and the R1S. And in sharp contrast to that in R2, we have a very limited set of build combinations that we're building today, which is all embedded in the launch addition. Even within that, we've limited the number of color combinations. And so that was not an accident. That was highly intentional to facilitate a smoother and faster ramp and that the decisions that we took around what combination of features we would have in this launch addition was also something we thought about a lot.

And as I said before, we had a whole bunch of our own expectations around what take rate would be to convert a reservation into an order for the launch addition. And we've been pleasantly surprised that, that conversion from reservation to order has been notably higher than what we expected, which is great news. And that doesn't, of course, mean we're not going to launch those other variants. So for any customers that are waiting for a standard trim or a premium trim, don't worry, we're on it. But those will be coming in the early part of 2027.

Your next question will come from Andrew Percoco with Morgan Stanley.

Can you guys hear me?

Yes.

Awesome. Well, I guess before I get to my question, I do want to say I was out of the 57,000 demo drives that you guys talked about. So as a potential future customer, I just have to give you guys some kudos here on what you've done with the R2. And just recognize that upfront. But I guess just to get to my question, we talked a lot about autonomy already, but I'm just kind of curious how you're thinking about the pricing side and potential uplift to ARPU as you go to point-to-point and then eyes off in 2027 and how you're kind of thinking about enacting those pricing changes over time.

We certainly believe there's opportunity to grow pricing for our autonomous capabilities as we add those capabilities, but we believe very deeply that, that needs to be earned through the expansion of the feature set and expansion of the of the capability.

I'd say the other element here, which we have to be cognizant of is just the market dynamics around this. And it's -- I think it's going to be a very quite fluid around how some of these features get priced. As I said, we're going through an inflection point where customers are going from this being a nice-to-have feature in terms of advanced autonomous capabilities to a must-have feature. There's a growing awareness around the capability set. And I really believe the transition to Level 3 capabilities where you can go hands-off, eyes off makes it much easier to implement or to use the feature. It's much easier to understand. Early adopters, it's fine to say you have to be hands-on -- or eyes on the road, but you're going to be hands off.

But for the broad full spectrum of customers, it'd be much easier to just say the car will drive itself. You don't have to look at the road. You don't have to touch the wheel. And we think when that happens, it will make it far easier to explain, far easier to see mass scale adoption well beyond just early adopters. And with that, the desire and the expectation for these features to be in technology forward vehicles. And so ultimately, what the pricing potential is there is something that we're actively spending a lot of time ourselves thinking about. But ultimately, we believe it represents an opportunity to expand pricing beyond where we are today.

Got it. Okay. That makes a lot of sense. And then I was just hoping you can talk a little bit about this new California rebate program. In my understanding, it does require some level of OEM buy-in. So I'm just kind of curious if that's something you're planning on participating in and whether or not you've started to see some incremental demand from that program.

Sure. We're really happy with the California EV incentive program. And consistent with our mission, we're really excited about the opportunity to attract new first-time EV owners into the Rivian community. So for the program, it's $3,500 for new vehicles and $1,750 for used vehicles with the OEM and the State of California splitting the share of those incentives. There are caps by OEM. And so there will be limitations in terms of the full population set that can take advantage of some of these incentives.

Your next question will come from Chris Pierce with Needham.

You've talked a lot about conversion. I guess I'd love to hear the flip side of that. Like what are customers that aren't converting telling you? Are they -- they put in a reservation a long time ago. They're waiting for a different variants. I guess, I'd just like to hear about people maybe falling out.

Well, thanks, Chris. Yes, of course, there's a long list of reasons that 1 may decide not to convert. But the biggest is they're waiting for a different build combination than what we have in a launch package. So that's by far and away the most common reason for someone to not convert a reservation into an order.

But as I said, recognizing that conversion rates are higher, that's been something we've had to dynamically really watch super carefully because we're providing delivery windows to our launch tradition customers and that requires us to make projections around what the conversion rates will be. Ultimately, when you think about working through the backlog of reservations. And so the way that we've tried to provide guidance on this is representative of thinking about those conversions, but ultimately beginning the delivery of those different trims in 2027.

Okay. And then I think you mentioned you might have another Autonomy in AI day this year. Could this be -- I mean it's been a while since the IPO and long-term financial targets were put out. Could this be something where things have changed for there for worse, for the better now? Like could we see updated long-term targets for people to sort of anchor around?

For the Autonomy in the AI Day, we'll be more focused on the technology road map. We're working over the coming quarters to think about the right time to set up a broader-based Investor Day as well that would be more focused on longer-term financial targets. But this one, RJ, I don't know if you want to provide a little bit of context of some of the updates or plans for the day itself.

Yes. We had our first Autonomy AI Day last year, and we pulled back on a lot of the technology development that had been underway for quite some time with in Rivian. So that was our in-house silicon program, the road map to Level 4 capabilities. And subsequent to that, we've now announced a few important things, one of which is the deal that we've done with Uber for our robotaxi program, which, as I said earlier, is going to be available in for targeting for that to be available in 2028. But there's a lot of other details that we'd like to provide around what that road map looks like. We'd like to show the progress that's happened on the vehicle to a what's in customer vehicles today, but actually from a technical point of view on the path to Level 4.

And of course, our objective is to provide better visibility and therefore, rethink a higher level of confidence around what those targets -- around achieving those starts.

Your next question will come from Philippe Houchois with Jefferies.

Yes. My first question is on the relationship with Uber. We've seen some tension between Uber and Waymo and their cooperation. And of course, there's a bit of uncertainty about the financial stability of Lucid and the fact that Lucid work with neuro. So we started as integrated a solution as what you offer in the parts with Uber. So I'm just wondering how much of a upside to your relationship there is, maybe not in the timing, but in the magnitude of what you could be working with Uber in terms of offering ruble taxes?

Well, a core element of our approach is a strongly-held belief that vertically integrating the vehicle and the technology stack together allows for meaningful, both cost advantages and technical simplifications in terms of the vehicle architecture and autonomy hardware architecture. And so that means how sensors are integrated into the vehicle, the redundancies built around the controls, the way that we've architected the compute platform in the inference in our case because we brought silicon in-house, we have a very low-cost way to achieve very high levels of inference within the vehicle. And so we believe that's a technical approach we've taken, from a hardware point of view, creates a structural cost structure, I should say, that's highly advantaged for delivering Level 4.

Along with that, we believe that the fact that we're building a neural net-based approach, an approach that's an end-to-end trained model with what we call LDM, or large driving model, means that the progress that we make on the consumer platform. So that's a point-to-point hands-off eyes off Level 3, ultimately getting into Level 4. That LDM workflow is identical across all those different steps. So that staircase up to Level 4 is not -- is one work stream. It's not multiple work streams where we have a Level 4 platform and a Level 2 platform, but rather there's one LDM, one end-to-end train model. And the vehicles will have slightly different sensor topologies to cover corner cases differently between, let's say, a Level 3 and Level 4 vehicle.

And along with that slightly different inference capabilities built around RAP1 but from a platform, from a model point of view, it's the same. And so that was a key shift for us. And I think it's a key shift for the industry in terms of how we approach autonomy that historically prior to, let's say, 2022, 2023, the technology topologies used for Level 4 were very different than the topologies used for Level 2. And we've seen that merge with the approach of using transformer-based encoding and the end-to-end models that is really to the advantage of our architecture to the advantage of the data flywheel that exists through the deployed fleet that we have.

But specifically on Uber, you think there's a big opportunity for you? Because I mean, when you announced the partnership, there was -- you were one of several partners that uberwas working with. Is there an opportunity for Rivian to be actually a much bigger partner of Uber in Autonomy?

Well, I'd say the working relationship with Uber has been fantastic from the top of the organization through the working levels of the organization. So the engagement from Dara in terms of the importance of this relationship, along with the working level teams, it's been very positive we're working towards our next milestone at the end of this year and deploying in a couple of cities in 2028 and then rapidly growing following that to a large number of cities. And so this is we're quite encouraged by the relationship and the partnership.

But ultimately, I do believe that the focus we have and the focus, I think, appropriately where we're placing our energy is making sure that the technology is ready. The vehicle is capable of delivering Level 4 performance and doing it in a thoughtful way in terms of really clean execution at the vehicle level. And really cost-effective ways of delivering that capability.

Right. And then maybe more for Claire about the ramp and -- I think you still haven't told us exactly how many R2s you delivered. It's just maybe your handful. But you had warned us about the risk to gross margin as you ramp up and in fact, I know there was a bit of an outside help in Q2 from IEEPA, from Z income. But the dip in the gross margin has actually not really happened. And now you're warning us that they could be in Q3. I'm just wondering, is it because the ramp really hasn't happened that much on R2 in the second quarter, or has it actually happened more in a smoother manner than you feared. And of course, you're cautious about Q3, but we could have a positive surprise in the Q3 gross margin as well.

As we think about the second quarter, you can think about us converting from a preproduction environment to our start of production and then start of our cost structure hitting our cost of goods sold for roughly 2/3 of the quarter. So it was not a full quarter's impact. And so as we look ahead to the third quarter, we'll have a full quarter's impact of the ramp. We'll also, as RJ noted, had the introduction of a second shift of operation and so some additional labor costs associated with the ramp as well.

And then in the fourth quarter, that's where you'll start to see the benefits of scaling production, the fixed cost leverage, both across R2 as well as the broader plant. And then with that maybe also invite Javier to jump in as well, just as we think about driving towards manufacturing efficiencies and executing the ramp in the second half.

Yes. Thank you, Claire. Indeed, there's a big focus -- laser-focused on the teams in normal to ramp up this first shift. That's what we started in quarter -- and we continue in quarter 3 with the preparation of R2. We are not expecting any material contribution to volumes with the second shift in this quarter, but we will see it in quarter 4.

The capability of the teams has very much increased compared to the launch of R1. And that's a big advantage. I mean the winning teams are the ones that are capable to solve problems at the speed. And that's what we are doing now, focusing on training of the members in the far lines, training the team, the automation and getting the output higher and higher as we ramp. And then, of course, focusing on the supply chain that is always not surprised. I believe many launches in my life and always the supply chain is a part that requires special attention. So we are focusing on a handful of suppliers with boots on the ground and with a great example of collaboration where the suppliers are really willing to get our support and we are willing to give them. So that's the focus in normal today.

This concludes the Q&A section of the call. I would now like to turn the call back over to RJ Scaringe for closing remarks.

Thanks, everybody, for joining us on today's call. Hopefully, you're observing a lot of excitement from the team here around R2. That excitement is echoing the excitement that we're seeing from consumers and it's openly been something that we as a business have been building towards for a while. The ramp that we're going to see in the second half of the year, so Q3 into Q4, as Javier just described with Q4 being a really important element where we start to see the benefits of our second shift online. That's a major focus for us. And our operational teams, our engineering teams, our customer-facing teams are all very much aligned to be focused on delivering on that ramp.

And as you also heard, we're continuing to focus on our technology road map. So that's continued progress on our software-defined architecture. And as we really think of it as our AI defined architecture, so vehicles that have a growing set of features and capabilities that growing the features and capabilities allows not only the vehicle to create a better customer experience, but exists within our ecosystem more efficiently.

And then, of course, the enormous emphasis that we have on our self-driving platforms, our self-driving platform and growing capability over the course of this year with the introduction of point-to-point late this year and then ultimately expanding that in 2027 and 2028 to ultimately get to Level 4. And so this is the exciting time for us in the business. We see it as an inflection point for us as a business as well, both in terms of our path to profitability, but also in terms of the scale and scope of the business given the mass market nature of R2.

Once again, thanks, everyone, for joining today's call and look forward to future discussions.

Thanks, everyone, for joining us for the next session. Very pleased to have with us from Rivian, Claire McDonough, CFO.

So just to get started, Claire, I mean, a couple of things we want to sort of really touch on here. But I think first and foremost is the upcoming R2, right, which I think everyone is sort of pretty excited about. Maybe you can sort of just tell us about how the preparations for that vehicle are going. I think it start -- officially sort of maybe opening things up to configuration sort of in the coming weeks. So where we are in the process and sort of how investors should sort of think about the cadence of R2 over the balance of the year?

Sure. Well, first off, Joe, thanks so much for having us. For those of you in the audience that haven't seen it already, we do have an R2.

Sorry, I should have mentioned that. Yes.

So definitely the better than listening to me is to go and experience the vehicle directly yourself, and you'll understand what we're going to talk about a little bit more today in terms of the intentionality in the design, the technology and the utility and performance that gets unlocked with the smaller and more affordable package that we've brought to market with R2. So on June 9, we'll kick off the first deliveries externally. So we've been making deliveries to employees and spend a lot of time working through the overall validation process before we start external deliveries on June 9.

And then also on that date, we'll also kick off having test drives across our 100 different service and sales locations across the country as well. So really excited to get more customers behind the wheel of the R2 and have them have the opportunity to experience it for themselves.

Okay. And I think in terms of sort of your guidance, right, you mentioned about 9,000 to 11,000 deliveries -- total Rivian, I guess, for the second quarter. But obviously, when you look at what you did in the first quarter, what you expect to do for the second quarter, there's a big ramp-up sort of here in the back of the year, which is to be expected, obviously, with R2. The sense I get is you feel pretty comfortable about demand, although any sort of commentary there would be appreciated. I guess what's sort of the gating factor for sort of where you sort of really think you can fit in? Is it just sort of like the pace of the ramp and sort of how the supply chain comes along? Or maybe just sort of help us understand the cadence of the year and what could go better or worse?

As you mentioned, our overall guidance, we delivered just over 10,000 units in the first quarter, 10,000 units is roughly the midpoint of our range for the second quarter. So that implies an average of about just over 22,000 units a quarter in Q3 and Q4. And with the cadence of the ramp, you'll certainly see more of those shifting towards the fourth quarter relative to the third quarter. As we think about the overall execution of the ramp, it's heavily going to be supply chain driven.

Our teams have spent a lot of time with each and every one of our suppliers and while I wish there's more flexibility in the automotive industry, you can only ramp as quickly as your weakest supplier. And so our teams are on the ground to validate, assess the quality, the manufacturing environment that each of our supplier partners are executing against today to ensure all of them can work as part of a broader orchestra as we approach our overall ramp for the back half.

One of the things I noticed about sort of the -- what appears to be sort of the launch plans and maybe this is sort of the lesson learned from the R1 experience, right, is a little bit of simplification of an offering, I'd say, to start, right? But I think that also probably also helps with keeping future demand unlocks going forward as you sort of unveil new trims and variants and offerings. So maybe you could just sort of talk a little bit about some of the thinking there. Is it really just sort of a risk and execution mitigation strategy to sort of start very almost homogenous and then sort of offer more and more variants over time?

It's part of an overall strategy of how do you go slow to go fast. And in doing so, we've cultivated a smaller assortment of our launch variant product, which is our performance variant, which has a number of additional bundled features included as part of it that we'll launch with. And then over time, in the back half of this year, we'll introduce our premium variant. But as we think about the order backlog of R2s that we have, we're just inviting customers that are interested in configuring their launch edition version of the vehicle to configure that, order that.

So we can also shrink down the time line between order and delivery and better manage expectations within the order bank of customers as well. So on the night, we'll unlock that for a select handful of customers, will be invited to configure their very first.

So you invite people who have orders to configure basically, not -- you won't -- someone from the general public won't be able to go on the site and sort of configure a vehicle.

Right. So we'll go wave by wave through the orders that we have on hand. And then the other element of it is we're going to have a data sample from customers that may have ordered more recently. It will help us assess the conversion rates across the overall cohort as we are doing our modeling and work associated with the ramp and deployment of the overall delivery execution as well.

What can you say about, I guess, the launch edition? Is that sort of -- is it sold out? And will it remain as sort of a certain variant -- high-performance variant over time, just called something different? Or how should we think about that?

So you can think about the Launch Edition as being our performance variant, but with -- it will have lifetime autonomy in it. It has a tow package included as part of it. So there are certainly different elements of what in the future will become added options that are included as part of that Launch Edition bundle.

Okay. On the R1, which I know the focus is on the R2. But I guess like the EV market, particularly in the U.S., which is most relevant for you guys, obviously, -- it's been through a lot of sort of the crossroads. I guess I'm curious if you could sort of reflect a little bit on how you think performance of that vehicle has held up in what's obviously been a sort of changing and dynamic backdrop for electric vehicles broadly. And maybe also a little bit of color on sort of how you're sort of seeing the demand and the order intake for that vehicle currently and what you sort of expect as R2 begins to ramp and also sort of obviously broaden out into additional trim levels?

For the R1, it still to the day is the best-selling premium electric SUV over $70,000. And in the state of California as well as the best-selling premium SUV over $70,000, ICE or EV included as part of the underlying denominator of vehicles. It's also a vehicle that established the Rivian brand to the market. It has a tremendous amount of fans, following, consumer reports as a study on intent to repurchase and the Rivian vehicles are 14 points above the next best player in market, really being a class-leading program for us.

But it is a much smaller addressable market as we think about vehicles in the $70,000 to north of $100,000 arena. which is where the game changes as we look to R2 and the opportunity to unlock a much broader addressable market of customers that can afford a $45,000 to $57,000 vehicle, and we're really excited to tap into that demand and the interest that we've seen in market for R2.

I guess to round out the current -- the portfolio as well, let's talk about the van and the relationship with Amazon. I mean, you do disclose every quarter in the Q, the sales to Amazon, it looks like it's been pretty strong, implying pretty decent RCV deliveries. Maybe just like mark to market on sort of the status of that relationship and what Amazon is telling you in terms of how they continue to want to sort of try to electrify their fleet.

Amazon has a really strong relationship with Rivian. And part of the beauty of the overall offering of electric delivery vans for them is the total cost of ownership advantages that they provide. And Amazon has taken years to build out the charging infrastructure to support the electrification of their fleet.

And so it's been really interesting to be sort of their partner on this journey as they've been able to work through adding power to some of their DSP sites to prepare for the electrification of their fleet over time. We're now in a position where we're seeing them increase their volumes on a year-over-year basis. And so we saw that certainly in the first quarter, and we anticipate there to be strong demand from Amazon throughout the course of this year.

And is that -- are there lessons learned there for sort of broader electric van opportunities? I know you sort of had some pilot programs with some other customers. But it seems like, again, maybe the pain point in why Amazon went a little bit slower earlier on was what you just mentioned to alluded to the infrastructure sort of build-out. So what are you seeing from other customers here as they -- in those pilot programs in terms of building out a support structure to support the vans?

I'd characterize there being interest and curiosity to understand what does it take to make the transition. And as you look at Amazon, which is the largest fleet operator, they see all of the benefits that the transition makes. But it certainly is challenging as you think about the investments in infrastructure, the lead times to prepare a fleet to electrify. And that's the piece that I think both Amazon and Rivian appreciate and understand and why we wanted to be balanced as we think about the size of external customers beyond Amazon in the near term.

So are you ready to take Amazon and lessons learned there and almost as a case study as you sort of go and talk to other customers about what it takes, what's needed? Has that helped with that process at all?

It definitely helps with the overall process. I think the other piece is it takes some time to educate on the total cost of ownership. It's not just the -- if I'm working in procurement and I look at the upfront cost or even as I look at the underlying cost of fuel or power cost or energy costs that are going into the product and the comparable set. The piece that is critical is the uptime that the vans are able to deploy, the fact that we're using mobile service, the vast majority of the time to help support them. So their vehicles never leave the DSP. We're going to the Amazon DSPs to fix the vehicles in real time that helps again really reinforce the total cost of ownership and operational advantages of what we're able to bring to market.

I guess moving down the income statement a little bit just on your outlook for the year. You did reiterate your EBITDA guidance for the year, I think, somewhat encouragingly, right, because there's definitely been some inflation on some of metals and other inputs. And so -- and plus, I think I know there's some R&D sort of accelerated, I think you sort of talked more about '27, but I imagine some of that probably creeps into this year as well. So maybe you could just sort of talk a little bit about some of the puts and takes and what gave you the comfort to -- I know there's a range, but like what gave you the comfort to sort of be able to reiterate that range?

The overall comfort and the reiteration of our EBITDA guidance is driven by both some of the net positives that we've seen, whether those be IEEPA, tariff refunds, whether it's increasing our sales of regulatory credits beyond what we had originally planned for or modeled, the continued visibility that we have into the R2 launch and execution as that stands today. while also ingesting some of the rising raw material costs that we're seeing across the industry.

I don't think you quantified potential IEEPA because I think you did say there was none in the first quarter, but there might be some in the balance of the year that is embedded in that number. Is that fair?

That's fair.

And credits, is there an update there? Because I think your guidance there was to basically assume nothing.

Right. It was more that we were able to execute more credit sales in the first quarter. Nothing. We're still in the -- don't plan for anything in the back half.

Okay. And order of magnitude on a potential IEEPA benefit recovery, any color there?

In the tens of millions of dollars generally how we size it.

Okay. Let's move on to Georgia and DOE. And I think what really sort of just love to revisit some of your thinking and calculus in terms of the renegotiations that took place. Obviously, like you upsized that initial sort of capacity, you get the money earlier. But the other -- I guess, if I were to sort of say somewhat balanced on that, right, like you are maybe giving up some -- a larger amount of money over a longer period of time to sort of help fund a larger build-out. Like not to say you can't expand beyond $300,000 there over time, but it seems like at least now, that's going to require you to fund that more or sort of other sources. So internally, how did you sort of think about that striking that balance between those 2 things?

As we thought about the impact to the increase in the Georgia site, we looked at it through the lens of the cost efficiency on a unit of capacity that we're building into the bounty, Georgia, where you see a lot of efficiency as you go from 200,000 units to the incremental 300,000 units. And so as we were dialing in our planning and preparation for it, that was a highly attractive investment in the future that we could make. And we were able to also get the Department of Energy to increase the initial phase of the loan as part of our renegotiation efforts with them.

So we actually increased the principal proceeds by $600 million. And so it created this really hyper-efficient initial phase of Georgia that takes Rivian to meaningful scale with total capacity of 515,000 units, a path with that scale to free cash flow positive in the future. And so for us, that first phase is really critical as we think about rapidly accelerating our growth in capacity in the near term, which this work is pulling that forward for Rivian.

Yes. I'm glad you sort of tied those 2 concepts together because I mean, one of the other things you obviously mentioned that got some attention from the investment community is the free cash flow positive at sort of that, call it, plus or minus 500 million units. So in some respects, this is maybe that allows you to get there a little bit earlier, right, because you're putting more upfront. But I guess like just how would you, volume is obviously important to being able to generate cash. But what are some other factors or are there other things that Rivian can do in order to sort of try to get to cash flow positive, maybe even a little bit earlier or at a lower level than 500,000?

There are certainly levels, and I can't sit here today and predict the exact GPUs that we will need 3 years, 4 years down the line or the cost of that end market. But there are certainly areas of accelerated investment that we're making where there's opportunity for greater levels of efficiency that we see in sort of the current state of affairs within the business today that could help, I would say, flatten the curve or flatten the volumes required for Rivian to be free cash flow positive in the future. The other dynamic is that it's heavily dependent on how aggressive are we building the next phase of capacity expansion as well, which, again, is a key barometer. But ideally, we're in a position where there's a tremendous amount of demand available to Rivian, and we're accelerating towards capacity expansion to fulfill that demand overall.

Let's switch over to ADAS and autonomy, the other sort of exciting vector here for Rivian. And let's start with the universal hands-free product, which I know is evolving more to point-to-point, I think, still by later this year. So how can you -- what can you tell us about sort of that technology path and sort of the progress that the products make? I know we were in sort of some early versions earlier this year, but presumably, you've sort of seen some progress there as more data and better product out.

Right. Lots of progress being made behind the scenes, not yet released to customers. We have released our universal hands-free offering, which allows hands-free driving on 3.5 million miles of roadways across North America. But the key offering for consumers is truly adding in the overall point-to-point offering for them so that they can have peace in mind as they're navigating both in town, on highway for each and every one of their trips. Our teams have been making significant progress. I was in a development vehicle a couple of weeks back in San Francisco.

In an environment like that, or I'm sure if you were to do the same here in New York City, you see just so many different edge cases that are happening left and right. And it was great to see how our large driving model was handling quite a dynamic set of circumstances and environment as well as we continue to progress our efforts there. The other piece that's critically important to our autonomous road map for the future is the ramp and scaling of R2. And the R2 fleet provides so many more edge cases of data that we can take back, feed into our large driving model, feed into the data flywheel that we have at Rivian to make our model is that much better and the quality of the experience and safety associated with it.

Yes. I want to put a pin on that because I think that's also where maybe some of the Uber relationship can sort of tie in, in terms of sort of getting scale and data. But just before we sort of go there, pricing for the product, like I think you're starting it at right now $50 a month, maybe $2,500 to buy it. I think you've indicated in the past that clearly, like as more functionality comes into the system, like you might have to reevaluate the pricing. I'm curious how you sort of -- like is that sort of a little bit of like experimentation, almost A/B test? Like how do you sort of think -- how do you figure out sort of what the right level is and what the customer ability to pay for that functionality is?

I would say autonomy pricing has been one of the most heavily debated topics within Rivian over the years. The philosophy and approach that we took is we want there to be widespread adoption of our capabilities. That also helps feed our flywheel, our models, more drivers utilizing these features is better and will continue to help us accelerate our progress as well. And so we wanted to start with a very approachable level that lots of consumers could say yes to. And so that was really the genesis and approach that we took as we thought about the pricing dynamics.

Yes. You mentioned some of the development vehicles upfront. Are those sort of the next-generation with the Gen 3? So -- and that's still sort of on track for later this year as well. And there were some reports that like you might look to sort of bring some of the LiDAR capabilities in-house. Is there anything you could comment on that?

So I think that was more of an erroneous headline. No plans today to bring it in-house.

I'm glad you could clear the air there. All right. So let's talk about Uber, which is obviously also interesting, and you definitely get your vehicles out there. And I think it also, again, as you mentioned earlier, get the data, get the flywheel. But I am a little bit curious how you sort of think about that, right? Because presumably, at least the earlier ones, and maybe I'm wrong, but like will they still have a like almost a safety attendant in there? And I know it might need to be dependent on sort of the geography you sort of going in. And if so, like is that on you? Is that on Uber? Like -- or you're not going to sort of really open it up to the public until you can actually sort of take the safety driver out or the safety monitor out?

We've already started data capture collection in the launch markets that we're working towards in 2028. As we're sitting here today, so what you'll see is just a continuation of the progression with safety drivers effectively executing point-to-point and ultimately bridging that to an L4 capability in the future overall. So we'll certainly go with our own progress, but always trying to create value-added solutions along the way.

Okay. So by the time you're sort of at that sort of scale number that you sort of have put out with Uber, it's sort of driver out type of operations.

Right. So the expectation is the 2028 milestone timing that we talked about being in Miami and San Francisco, those are driver out milestones that we're executing towards.

And how do you think about -- right, so there's obviously sort of benefits from using a network like Uber in terms of leveraging their network and you get the benefit of the miles. I mean I do think, though, one of the things you maybe lose out a little bit on is sort of some element of the customer experience. Obviously, not sort of being in the car, but Uber as is widely seen like they are -- they plan to have multiple sort of AVs on their network. So what were some of the like strategic considerations in sort of putting your vehicles on that network? Because it almost seems like by trying to fragment the market, they're sort of trying to get rid of sort of differentiation of product and transportation on that front.

For us, the core consideration in partnering with Uber was to focus our investments and team on what we see as the highest impact area of robotaxi, which is focusing on the development of the driver. And so instead of trying to do everything all at once, being able to focus on that, which we think is great for robotaxi applications will also unlock consumer L4 opportunities, which we think is an enormous market in the future as well. That was our priority and focus. The other dimension, which is helpful as we're continuing to ramp up our capabilities and performance is the opportunity within the Uber application to occupy and grow your ODDs over time as well.

And so if you're starting up a service and you're just offering in-town rides in certain areas or locations to kick things off, that's a harder hurdle to get through versus within the Uber app, essentially Uber can direct rides that meet Rivian's ODD and then we can develop and grow as we continue to expand into all environments over time.

So -- but as the technology evolves and the -- let's say, like personally owned autonomy path and like the robotaxi autonomy path converge, and [indiscernible] sort of has always sort of talked about -- or he first started talking more about personally owned autonomy versus sort of like this robotaxi, right? So -- so how do you sort of think about, right, like as those converge down the road, someone owns an R2 and maybe they do want to be able to sort of monetize their vehicle while they're not using it and is the plan that they would also sort of be able to put it on an Uber or Uber-like network? Or down the road is -- are some elements of that transportation network company type infrastructure, something you would look to expand beyond Uber or even sort of handle internally?

The great element that Rivian has is because we're a direct-to-consumer business, we have a service network across the country. We have charging infrastructure across the country. And so we have a lot of the core tenants and attributes that could make Rivian successful in terms of the operation of large-scale deterministic fleets. And if you look back, RJ's original thesis in creating the partnership with Amazon was this very training and learning of how do you run and manage a very large fleet of vehicles for the evolution of a shared mobility type of future as well. So we're certainly today looking at lots of different options and see lots of potential given the assets that Rivian has and the capabilities that we have in-house.

Okay. One of the things I think that's become a lot more in focus is and that we've been focused on as well, and I think investors have also is looking at what the core competency of the business is and whether there are -- whether that provides sort of additional opportunities beyond that. Rivian, I think, has probably been sort of more on the, let's say, the leading edge of this, right, like you've designed your own chip, right? You obviously sort of designed your own electric architecture. I think you designed some of your own power electronics. You have also, which I mentioned to you earlier, I keep getting in my social media feed and I'm tempted to try to buy one of those electric bikes, they look pretty cool.

And then obviously, robotics as well, where I think you own 38% of -- right Yes, 35% Okay. So look, I guess what I'm getting at is, I think -- one of the ways you could look at Rivian is that you have a very sort of specialized set of employees and know-how that know the convergence of software with hardware. And that you've got a big task in front of you, especially sort of R2, but there are also other potential applications of the resources and capabilities you have. So that's a big question, but how do you sort of think about that internally? How do you sort of go through the process of weighing whether this is something you should pursue? Because like even with bikes and robotics, like you did, at some point, make that decision to sort of commit some resources to that even if they sort of eventually have been and further capitalized by others.

I think to your point, Joe, at Rivian's core, we're a technology company. We're a technology company that's operating in the mobility space. but we are a technology company at our core. And one of our core advantages is our ability to attract and recruit top talent across electrical hardware, across the semiconductor space, across software development space that I think uniquely positions us within the broader automotive and mobility sector as a whole. These are becoming, I would say, more and more attractive resources to the industry as we think about the advent and introduction of more physical AI.

The other core attribute that Rivian has is our data. And so whether that's the data in our manufacturing environment, which is rich for the likes of Mind Robotics to learn, develop and grow from, whether it's the data of our fleet of vehicles that is capturing and helping to inform our large driving model for autonomous driving. Those are really central to the evolution and acceleration that we're seeing in the broader universe of physical AI applications. And we're really excited about the rate and speed of growth and advancement that we're seeing in each and every one of those areas.

Maybe we could just sort of zeroing on the chip, for instance, right? Like I think like when you put that out, it showed some pretty impressive performance. Now I know that was sort of specifically built for your use case, which I would sort of say is visual large language, right? And -- but to your point on physical AI, that clearly has maybe other sort of applications. So is there an organic effort to sort of try to widen the aperture for what that chip could do or where it could be applicable or sort of more you got to sort of focus on your task at hand. If someone comes to you, you're willing to engage in a conversation if the economics and opportunity are right?

As we think about the chip capabilities, we see it not just in automotive types of applications, but we see it lending itself quite well also to the universe of robotics. And I'm really excited about the potential, not just for Rivian, but for technology licensing as a whole across many different dimensions of what we're building in-house as well.

Okay. Within automotive or outside of automotive or both? It can be both. Maybe just a -- let's sort of close here or approach a close on the Volkswagen joint venture. I know you have the winter testing done. It came with some -- meeting some thresholds that got you some capital in the door. Can you just remind us one, what's left to do in that venture? I think the ID1, which I believe is the first vehicle is sort of scheduled to launch at some point next year. And how does the sort of P&L sort of change once the output of that JV starts to sort of get into vehicles? Like will you start getting payments right away? Or is there some sort of like bank that was sort of built up from like prepayments, if you will, that you sort of got to be eaten through first?

As you think about the $5.8 billion joint venture, there's sort of still to come payment of just about $1.5 billion. A $1 billion of that is a nonrecourse loan that Rivian expects to receive in October that has no milestones associated with it. And then the final payment will be the earlier of the first vehicle, so the ID1 in this case, being commercialized or January 2028. So there are -- from a payment standpoint, actually no physical milestones left because of the time-related feature of that final payment for Rivian. But regardless, we're really excited about the evolution of our in-house electrical architecture software stack and the opportunity to scale that with multiple different Volkswagen Group programs. So we completed the winter testing for 3 different brands and 3 vehicle programs at Volkswagen Group this past winter and are excited to help them commercialize those vehicles over the course of the next year or 2.

Okay. And then again, as those vehicles get into production, how does that change your P&L?

And then as you think about the P&L change, the piece that's important to recognize is about $2 billion of the overall consideration is associated with consideration for background IP. So we've been recognizing that $2 billion from the close of the JV, and that will be recognized through the middle of 2028. And then beyond 2028, we'll flip to VW paying Rivian $100 million a year of just ongoing fees.

Okay. So it's not like a per -- license fee per...

It's not a per vehicle license fee because we're sharing in the joint -- the development costs jointly. So VW pays for 75% of all of the shared development work that occurs within the joint venture and then Rivian pays for 25% of it. But because of those splits and dynamics, there's not an ongoing per unit fee that occurs.

And just since we're somewhat on sort of the capital front also, just going back to the DOE loan, that's expected to start coming in next year as you start -- and is the way it just functionally works is you build out part of the facility and they almost pay -- they loan the money as you sort of go in terms of sort of what you spend? And is there any timing mismatch on that? Or how does that work?

Yes. So the -- you can think about the DOE loan is really almost like a project finance instrument. So Rivian will contribute the equity of the project into the new Horizon entity. And then as we continue to build out the site, there'll be draws from both the equity pool from Rivian and then we have an up to 80% loan-to-value from the DOE. So they'll be reimbursing us along the way for the CapEx that's deployed.

Okay. And so that -- and -- but that capital or that loan, I should say, that basically we should think about it from time from sort of when -- when it sort of starts in earnest, I guess, at some point next year until sort of SOP at Georgia?

Correct.

Okay. Perfect. Claire, thanks very much for joining us this year. Really appreciate it, and thanks for the great conversation.

Thank you. Appreciate it.

Energy mobility, my partner, Davis Sunderland here. I'm very happy to have Rivian and Claire McDonough, the CFO. Joining us. Thank you, Claire. I appreciate it. I know it's a busy time for you guys.

Maybe if we could just level set, just I think everyone knows who Rivian is. If we could just level set maybe where you guys are right now, like what are the big things that you're focused on?

First off, Ben, thanks for having us. It's a really exciting time at Rivian. So next week on June 9, we'll start external deliveries, test drives of the R2. So if you go across North America to our sales showrooms to our service centers, you'll be -- have the opportunity to experience the product directly. We've been, over the course of the last handful of months going on and hosting R2 block parties at a number of sites across the country. We've seen thousands of people come wait 45 minutes an hour just to sit in the vehicle and experience it but not yet drive it.

And so we're really excited to have people and our future potential customers behind the wheel of R2. And this is just a tremendous amount of effort, energy that went into the engineering design of the product, the design of the technologies that feed the R2. And importantly, now the ongoing operations and scaling of our R2 production in normal Illinois.

Maybe just another context question, I guess, just to guide our conversation just about the broader EV landscape. And where Rivian kind of sits within that because we went through a period of excitement and then we kind of peeled back and had tariffs and losing tax credits, and now we're somewhere in between. And I wonder if you could just talk a bit about what you're seeing in the broad market, whether it be from the perspective of supply chain or consumer demand or however you'd like to take it?

Sure, as you take a step back, we always had the intention of building the best product. So RJ, our Founder and CEO, he wasn't out to create the best electric vehicle. He was ready to create the best vehicle full stop. And that was the mentality and approach that we took as we designed and developed the R1T and the R1S and our launch products.

And it certainly testament as we think about the opportunity to significantly reduce the cost of the R2, but create a vehicle that is incredibly fun to drive. I can't wait for many of you to get behind the wheel of an R2 to experience the thoughtfulness and intentionality that goes into thousands, if not millions of decisions in which we can design a product that keeps what makes Rivian essentially a Rivian, which is the adventurous spirit, the capability, utility and performance that customers have known to expect when they think about a Rivian, but doing that through a lot of underlying design and technology efficiencies that the R2 is now leveraging in market as well.

So maybe in normal, can you talk about just what production will be like there just from the R1, the EDV and R2 to start out and then we go from there.

Sure. Within our normal Illinois manufacturing facility, we have 215,000 units of total potential capacity split between three vehicle programs. So we've engineered the new body shop and general assembly for R2 support 155,000 units of annual capacity and then have a level of flexibility as we think about the R1 as well as the commercial van production that can complement the volume that we anticipate building at scale with R2.

We're starting first with just a single shift of operations. And we're starting with our launch products. So we have a more limited assortment of vehicles that we'll be building out of the gate that helps us streamline the supply base, streamline the efficiency as we continue to increase our volumes each week and ultimately scale up to full production.

I want to ask about Georgia capacity, too, but you brought up an interesting point about supply base and streamlining that. And wondering if you could just talk a bit about similar parts or similar technology or I guess, similar things that you can leverage in the R2 that were, I guess, preexisting from the R1 already?

One of the core foundations for R2 was Rivian's transition from our Gen 1 R1 to our Gen 2. And as you think about the core technology foundations of R2 in our electrical architecture, our software stack, our autonomous capabilities in the vehicle or even our propulsion with our drive units, again, all of them are foundational based off of the Gen 2 R1. They just take another step in terms of the efficiency and development from that point in time. So we often talk about the technology evolution from Gen 1 to Gen 2 as being even greater than the shift or leak that as we go from the Gen 2 R1 to the R2.

So one thing that I think investors are focused on is the R1 not being profitable. Can you talk about what R2 in the platform and how that changes?

Sure. For R2, we focused on both commercial negotiations with the supply base where we had a very different starting off point than we had when we started out with R1 and we're a preproduction company negotiating with suppliers. We also had the advantage of having just recently announced the Volkswagen $5.8 billion joint venture at the time when we started sourcing parts for R2. And so that was a real complement of validation for Rivian's future in the space and the opportunity for us to jointly source and procure with the Volkswagen Group, the electronic control units and common parts that come out of our joint venture together that ultimately drives another step change in terms of affordability and sourcing leverage for Rivian as well.

And then beyond the sourcing side, a lot of it is driven by engineering design efficiencies in the vehicle. So a couple of good examples are the suspension in R2, which is 72% less than -- or sorry, 72% of the cost of -- sorry, less than the R1. And that's driven by us evolving from an electrohydraulic dampening -- active dampening system to the MacPherson strut potential. So we still have tremendous off-road driving dynamics with the current suspension system in the R2, but are doing it in a much more cost-efficient manner as we think about the vehicle's execution.

For another great example that I love is as you look at the rear door frame of the R2. In the R1, we have a division bar. So you have multiple parts of glass in the vehicle. The R2 is able to design the door frame with no [ div ] bar included. It's actually aesthetically even more appealing with just a seamless single piece of glass in that door. And again, we're able to reduce 50% of the cost of the rear door as part of that specific example as well.

You mentioned in there, the partnership with Volkswagen in the $5.8 billion JV. One of many important relationships you guys have announced company life to date. I'm sure we'll dig more into each of them. But maybe just to level set on Amazon, now Uber most recently and then the Volkswagen partnership. If you could just give an overview on where those stand, that would be helpful.

Sure. Rivian has always wanted to be seen as a great partner for the industry. And our first partnership was with Amazon and centered around our electric delivery van. And it's been great to see Amazon scaling out their infrastructure from charging at their sites. That they're now in a position to increase the volume of EDVs that they're purchasing from Rivian, which is a key tailwind that we talked a little bit about on our prior earnings call as well that we'll continue to see over the course of of 2026.

As we think about the Volkswagen, joint venture was fantastic to execute and complete our winter test milestone, which was testing the electrical architecture and software across Volkswagen Group products, Audi and Scout products, and this past winter in Sweden and seeing the ongoing development and now commercialization, just about 1.5 years following the closing of the JV, which really demonstrates the extensibility of the technology that we're building and its application to many form factors in the future as well.

And then finally, our most recent partnership with Uber, which is centered around robotaxi deployments, the investments from Uber are allowing us to accelerate the development of our L4 capabilities as an organization that we see not only opening up new business models, such as the robotaxi deployment, where Uber is purchasing 10,000 units initially and up to 50,000 units in totality and each of those units with an associated software subscription for the autonomous driving from Rivian.

We also see it as a key catalyst for the end state of consumers where autonomy is going to be a more and more important part of the decision-making process that each of you make as you think about your next vehicle, the headroom, the performance, the capabilities that it can provide and the convenience that it can give each and every one of us our most valuable asset, which is our time back.

Just maybe going back to the VW partnership. I think that when you guys did it, at least I looked at as a way to monetize technology -- but I think there's probably more to that than that. Can you just talk about how it can benefit you from the P&L? You talked about from cost, but how it else?

So there's multiple vectors of savings and efficiencies as we think about the Volkswagen partnership. First and foremost, as we talked about the technology development, it allows Rivian to become a more global player faster with a broader R&D budget associated with localization efforts across the world and especially as we think about Volkswagen's position in Europe.

And so first and foremost, we think about it as the R&D pool of investments that can help Rivian continue to be at the leading edge of technology advancement and deployment within the vehicles. The shared development work is also split. 75% Volkswagen is paying for it and 25% Rivian is paying for it. And so there's inherent and out of the gates through 2028 R&D savings to Rivian where previously we're putting 100% of the bill of our team, whereas now they're focused on development for both parent companies as part of the joint venture.

We've talked a little bit about the sourcing leverage and opportunity, and we see there being potential to have additional adjacencies of sourcing leverage in the future together with them as well. And then the capital as a whole as we think about the $5.8 billion of total consideration.

And just as an offshoot from that, you recently got another one of those milestones of that capital stack that you guys will benefit from. What are the remaining steps to unlock the remainder of that?

So we anticipate receiving another $1 billion of nonrecourse debt in October. There's no milestone associated with the nonrecourse debt. And then there's an additional $460 million that will be paid when the very first Volkswagen Group programs are out in the wild commercially as well. So that will be the next payment, which we anticipate happening in '27.

So maybe let's talk about Uber and how that relationship started and the details about it.

Sure. As you all know, we've had autonomous ambitions for quite some time now as an organization. And we are working through the right cadence and timing and trade-offs as we thought about our own accelerated development towards L3 and then ultimately, L4 capabilities.

We spoke a little bit about this at our AI Day in December of this year. And as we continue to explore the importance of autonomous driving and the potential value creation, margin expansion and the fact that we see Rivian as one of just a few players that has the core ingredients of being able to be quite successful in this arena. We made the election to say let's go faster towards this future. But let's find a partner, as we found in Uber, who can help with some of the demand aggregation can help with some of the fleet management capabilities while we can really focus our efforts on building what we think is the most important part of the equation, which is the driver.

Maybe going back to that AI Day and talking a bit about that Rivian driver. And when you guys unveiled the Rivian autonomous processor, could you just talk a bit about the decision of why to make your own chip, what the competitive advantages may be? And then obviously, that's a supply chain that's very tight right now. If you could just talk about the relationships you have there to manufacture, that would be helpful, too.

In December, we revealed what we call RAP1, which is our very first Rivian Autonomous Processor. That will go into our R2s end of this year. And the core advantage for building a chip like this in-house was both speed. So our ability for our software teams to work on early releases of the product as we've seen over the course of the last year in the future development of the software stack and middleware that sits on top of the in-house chip.

It also allows us to really distill down to what matters most. And so if you think about an off-the-shelf offering, the shelf offering needs to suffice and build capabilities for a broad set of use cases beyond just the determinant of the physical AI in a camera-based application, which Rivian is clearly focused on. And so you get to a higher level of utilization of the core technology that you're developing because you're designing it to be a bespoke to your end use case. And the last piece is cost. The semiconductor universe of players certainly has great margins associated with it as well. And so the combination of speed, performance and cost really comes together as we think about the opportunity set for us on a go-forward basis.

I guess you get this question, we can go back to that. But are there other pieces that you want to monetize of your technology? Because it seems like people ask you that after VW and now you have Uber and is there anything else?

There certainly is a number of ways that we could continue to monetize the hardware and software for autonomous driving with other OEMs. Certainly, that is also true as we think about the electrical architecture and software that we've designed, developed in the JV and the proof point of having it situated in a number of different Volkswagen Group brands and products in market allows it to be a lighter lift as you think about what does it take to put it into an incremental vehicle program with another OEM as well. So we certainly have designed our technology stack for sizable scale in Rivian's long term, but certainly also see opportunities for partnership via licensing or monetization as well.

Continuing the Uber piece. And I guess going back to what you were talking about with Volkswagen and sharing some of the development expenses, could you just talk a bit about maybe some shared autonomy expenses or, I guess, ultimately, robotaxi expenses and how they might be split between you and Uber in the partnership?

So within the Uber partnership, Rivian is responsible for the development of the core technologies. Uber is investing in Rivian. And the way I think about it is we sort of shoehorn those Uber investments in Rivian to help offset the acceleration of our efforts in our L4 development road map within the business. But they are not sharing in the cost and the development directly.

So R2 next week, I'm coming out to California. One great thing from my personal perspective is that you kept the brand similar. How big do you think the market is for the R2?

We think the market is enormous for R2. So if you look at the broader market across U.S., EMEA, the 5-passenger SUV and crossover space is the largest, fastest-growing and most attractive space to play in and R2 really sits at the epicenter of it as well. So we see this as being a product that we hope a lot of customers that may have been on the edge of buying their very first EV now have the opportunity, a form factor that shares the adventurous spirit of the Rivian brand for them to delve into their first EV as well.

I echo Ben's sentiment. I think the R2 is very cool. And I was very surprised at the R2 unveil event when you also unveiled R3. Could you also talk about R3, how that fits into the platform and when that might be coming after R2?

Sure. We get questions about all the time. And part of the reveal was when you're building vehicles that will be situated on the same common platform. You want to make sure that the design elements are done at a similar point in time. So for our Georgia facility, we'll be building units across the midsized platform. So that will house R2, R2 robotaxi for Uber, for example, we'll house R3 and additional variants in the future as well. So we're right now in the process of digging the foundations of our stamping press building in Georgia and are excited about the continued progress on the creation and erection of that site.

So maybe you could give us the sizing of Georgia. You have a loan guarantee you guys upped the size of it, I think. So just maybe if you could level set us on that.

Sure. So we recently increased the initial first phase of Georgia to be 300,000 units. And associated with that initial phase, have a $4.5 billion Department of Energy loan. We did modify from an original plan that was going to have two phases of 200,000 units and a slightly larger DOE loan associated with it.

But what was the advantage for Rivian was we were actually able to increase the size of the initial loan and are able to create 300,000 units of capacity alongside it that is more cost effective for Rivian as you think about what is the CapEx deployment per unit of capacity that we're achieving and then having it up to 80% loan to value to help finance it with U.S. treasury rate capital, which is fantastic for Rivian's cost of capital.

Going back to the profitability discussion around the R1 and on the R2, broadly speaking, could you just talk through some of the biggest buckets for margin improvement in either of the vehicle platforms?

In R1 or R2?

Either, yes.

In R1 or R2. Either, yes. The areas of margin improvement are -- we gave a couple of examples on R2 with the suspension system, the door. Another couple of examples there are the battery pack that we're using, which is also -- it's a structural pack. So it's also the floor of the vehicle. We've also moved to some large die castings in the vehicle as well that again, simplifies the manufacturing process for R2 and also reduces cost as well. R2 is also about 2,000 pounds lighter than R1. So that helps us drive range and efficiency. And then we've designed and developed what we call our Maximus drive unit, which again removes significant cost from the program while also creating a really fun 660-horsepower driving experience that I think will make most of you smile as you hit the accelerator as well.

When I -- when I said how big is the market, you said EMEA and you guys haven't talked about selling too much into Europe or the Middle East. So what's the strategy there?

So we definitely designed R2 to be a global platform. And so we do anticipate in the future, we'll start here in North America, where we have a robust service infrastructure, sales infrastructure. But as we scale a product over time, we certainly have ambitions to take it more global as well.

And you guys have your stores or spaces. Will you do that in Europe? Is that the thought?

Yes. Spaces in Europe as well.

So maybe thinking about other just potential hurdles, not to the R2 but to EV adoption one that comes up frequently is charging. Could you talk just a bit about the Rivian adventure network, where this stands and what the investment plan is for charging infrastructure going forward?

The Rivian Adventure network today has just under 1,000 chargers across about 145 locations in North America. We've continued to grow the network over time. It's also a network that we've opened up. So it's going to be in the evolution of now being more NAX native as a network, but having CCS capabilities for vehicles as well.

And in opening up the network, we're actually seeing broad adoption from all different types of EV owner. So it's a great touch point for new potential customers to have into the Rivian brand and ecosystem, and they really appreciate the high level of uptime and reliability that we have in the network as well. We've definitely been more modest in terms of the funding road map and have relied heavily to date on government grants and funding this helped augment the capital investments that we've made in the site, but certainly have large ambitions to continue to grow out the network over time.

And I think the Amazon relationship doesn't get talked about a lot. Could you just give us an update there and how they think about is it EDV and/or anything bigger than that?

With Amazon, we've designed from the ground up with their use case in mind for last mile delivery. And the great thing about the van is the total cost of ownership advantage that it provides Amazon. And as they think about the efficiency of not just the ownership experience, but the efficiency of utilization for their drivers and the execution of the deliveries that they have. We also are introducing a larger pack variant of the vehicle and an all-wheel drive variant as well so that they can start to use the EDV for more rural routes within their network, which is a big area of growth for Amazon. So we're excited about the road map for the future and to continue to scale and grow the bands for them.

I know we'll get to one on capital allocation and balance sheet before we close, but I want to separate this question on strategic priorities separate from capital priorities, just in thinking about R2 being front and center, recently announcing an autonomy partnership with Uber and seemingly so many things going on all at once. How do we think about what is the focus or what is the most near-term, I guess, obstacle for Rivian to tackle?

So I see them as very intertwined in that you can't have the R2 without the foundational technologies that power it. And you can't just have R2. R2 would not be the same vehicle that it is without those foundational technologies. And so as we think about the prioritization, it's really centered around the operational issuance of scaling R2, being able to get more customers behind the wheel, deliver on the ribbing experience at scale. While, at the same time, continuing to progress our technology road map, which is foundational to current products as well as the future products that we'll be bringing to market.

We can do one before capital allocation. You guys have invested in the robotics company. I think. Could you just talk about that?

Sure. So we saw -- and this sort of dovetails into the question on capital allocation and how do we think about prioritization. With the advent and growth that we're seeing in physical AI, we saw a huge opportunity to invest in robotics for industrial and manufacturing environments. And we recognize that as we think about the uses of capital in our core business that it was important for us to incubate and fund Mind Robotics, which we started last year with outside capital. And so RJ has been highly successful in raising additional capital for Mind Robotics, and we're really excited about the development road map and opportunity for Mind to partner back with Rivian as we think about the future of manufacturing.

And last one, just as we think about your cash runway and how we think about capital as you progress and ramp up the R2 and get to Georgia, how do we think about that?

Sure. As you think about the cash runway, we have a number of sources of capital. So we ended last quarter at Q1 with $5.4 billion of overall liquidity. As we look at this year, we anticipate receiving another $2.55 billion of capital between Uber and Volkswagen. $1.3 billion we already received sort of right after earnings time frame.

And then beyond that, there's another $700 million from Uber, another $460 million from VW and the $4.5 billion DOE loan, which puts us at sort of a total pro forma capital and liquidity of $13.6 billion. And beyond the capital that we've secured from these strategic partners in DOE, Rivian will continue to be opportunistic as we think about our road map for the future and ensuring that we're always maintaining a strong balance sheet.

Great. Thank you, Claire.

Thank you.

Good afternoon, and thank you for joining us for Rivian's First Quarter 2026 Earnings Call. Today, I'm joined by RJ Scaringe, our CEO and Founder; Claire McDonough, our Chief Financial Officer; and Javier Varela, our Chief Operations Officer.

Before we begin, matters discussed on this call, including comments and responses to questions, reflect management's views as of today. We will also be making statements related to our business, operations and financial performance that may be considered forward-looking statements under federal securities law. Such statements involve risks and uncertainties that could cause actual results to differ materially. These risks and uncertainties are described in our SEC filings and the earnings presentation we filed with the SEC today.

During this call, we will discuss both GAAP and non-GAAP financial measures. A reconciliation of historical non-GAAP to GAAP financial measures is provided in our earnings presentation and press release. Just before the earnings call, we posted our earnings presentation, which includes an overview of our progress over the recent months and replaces our shareholder letter. I encourage you to read it for additional details around some of the items we will cover on today's call.

Following our prepared remarks, we will be taking questions from sell-side analysts. In the interest of keeping our call to 1 hour, we would ask these analysts to limit any follow-on questions to one.

With that, I'll turn the call over to RJ.

Thanks, Chip. Good afternoon, everyone, and thanks for joining us for today's call. Last week, I was thrilled to celebrate the start of saleable R2 production with our team at our plant in Normal, Illinois. It's an exciting milestone in Rivian's history and the culmination of all the hard work and energy from so many people across the company. As I've said before, I believe the R2 will be a game changer for our customers and will be a key driver of our company's long-term growth and profitability.

In an American automotive marketplace starved for high-quality EV choice, I believe R2 is an attractively priced option sized for everyday ventures from school pickups to weekend trips that is targeting the very popular 5-passenger SUV and crossover segment.

With R2, we are taking our design, performance and technology and bringing it to a significantly broader audience without losing what makes Rivian unmistakably Rivian. We've started R2 deliveries to our employees, and I have to say I absolutely love having R2 as my daily driver. I could not be more excited to get this vehicle into the hands of lots of customers starting this spring.

In developing R2, our team relentlessly focused on achieving structural cost reductions while maintaining the desirability of the product. For R2, our bill of materials is expected to be approximately half of our R1 platform. For non-BOM cost of goods sold, we expect to see a reduction of more than 50%, resulting from a focus on design for manufacturing and leverage fixed cost efficiencies through higher production volumes. This is how we expect to profitably deliver R2 at an accessible price point at scale without compromising performance and utility customers love for Rivian.

Key design changes for R2 include part eliminations and reductions through the introduction of large die castings, a structural battery pack, a new highly efficient drive unit, the evolution of our next-generation electrical architecture, which removes miles of copper wire and the consolidation of our high-voltage electronics into a single enclosure. We are also seeing significant sourcing leverage relative to R1 across a variety of components.

Now as we begin to scale our operations in Normal with R2, we're very excited to partner with the U.S. Department of Energy to grow our manufacturing footprint in Georgia. R2 provides the opportunity to expand the Rivian brand to millions of drivers. As a result, we made the strategic decision to increase the production capacity for the first phase of our Georgia plant by 50%, bringing it to 300,000 units of annual production capacity for our midsized vehicle platform. This change is expected to boost cost efficiency while still providing significant room for future expansion in later phases and support thousands of jobs in Georgia as we grow American manufacturing and work to ensure the U.S. retains its leadership and innovation in technology and transportation. We remain on track for the production of our midsized vehicle platform to begin in Georgia in late 2028.

Turning to our technology road map. In March, we were excited to announce a new strategic partnership with Uber to accelerate our shared autonomous vehicle goals. In the not-too-distant future, I believe advanced autonomy capabilities will be a key differentiator for customers and the driver of market share. At the core of our third-generation autonomy hardware is the Rivian Autonomy Processor or RAP1. The development of our RAP1 chip is on track, and we are progressing well on validation and reliability testing. Our integrated approach allows our hardware team to rapidly iterate with our software team, and our autonomy feature development is progressing well, and we continue to expect to begin rolling out point-to-point capabilities by the end of the year.

Finally, in the coming weeks, we are excited to launch the Rivian Assistant on R1 and R2 vehicles. The Rivian Assistant is our new AI-powered voice assistant that is built to be a digital copilot with integration into the vehicle ecosystem and other external apps.

In closing, this quarter, our team has executed across many fronts, laying a strong foundation for the years ahead. As an American automotive technology company, we're building for a future that we believe will be fully electric, autonomous, and AI defined. With our category-defining brand, the launch of R2, which is our first mass market vehicle, vertically integrated and extensible technology and the direct-to-consumer sales model, I couldn't be more excited about the opportunity ahead for our customers and for our business.

With that, I'll pass the call over to Claire to discuss our financial results.

Thanks, RJ, and good afternoon, everyone. As RJ shared, the start of saleable R2 production and initial employee deliveries are a landmark moment for Rivian. By building R2 in Normal, we are strategically leveraging our existing manufacturing footprint in Illinois to drive greater fixed cost absorption across our entire vehicle portfolio. As discussed previously, R2 production is starting with a single shift operation, and we expect to scale to 2 shifts by the end of 2026 as we ramp towards our North Star target of profitably delivering 4,000 vehicles per week in normal. Delivering a strong 2026 exit rate for R2 production and deliveries is a key focus for our team as we believe it will directly translate into positive automotive gross profit for the business.

Turning to the results for the first quarter. As depicted on Slide 11 of the earnings presentation, our consolidated revenue in the first quarter was approximately $1.4 billion, an 11% increase over the same quarter last year. Consolidated gross profit was $119 million, and our gross margin was 9%. Gross profit included $122 million of depreciation and $27 million of stock-based compensation expense.

Adjusted EBITDA losses for the first quarter were $472 million, driven by our $119 million of gross profit and increased adjusted operating expenses as we prepare to scale R2 and invest in our autonomy road map. In the first quarter, we produced 10,236 vehicles and delivered 10,365 vehicles, which was the primary driver of our $908 million of automotive revenue.

Automotive gross profit loss was $62 million compared to $92 million of gross profit for the same quarter last year, primarily driven by the $100 million decrease in sales of automotive regulatory credits and lower production volumes, which resulted in a $45 million increase in depreciation and stock-based compensation expense combined. While current macro and geopolitical factors are creating added complexity, cost and uncertainty, our team continues to work hard to manage supply chain risk and offset elevated costs.

Our Software and Services segment reported another strong quarter as depicted on Slide 13. During the first quarter, the segment generated $473 million of revenue, a 49% year-over-year increase and $181 million of gross profit. $282 million or approximately 60% of Software and Services revenue was attributable to our joint venture with Volkswagen Group. We also experienced strong growth from remarketing and parts and service.

During the quarter, we also recognized $506 million gain in other income in our financials related to the Series A capital raise and related deconsolidation of Mind Robotics from our financial statements. We currently own approximately 38% of Mind Robotics on a shares outstanding basis.

Looking at our balance sheet, we ended the quarter with approximately $4.8 billion of cash, cash equivalents and short-term investments. With regard to our funding road map, in 2026, we expect to receive a total of $2.55 billion of capital from our strategic partners. Today, we received $1 billion from Volkswagen Group in exchange for equity following successful completion of the winter testing milestone by RV Tech. The testing program spans several months utilizing reference vehicles from the Volkswagen, Audi and Scout brands.

Later this quarter, we expect to receive $300 million from Uber in exchange for equity related to the signing of our partnership agreement, subject to certain conditions. And later this year, we expect to receive $1 billion in nonrecourse debt from Volkswagen Group and an additional $250 million from Uber in exchange for equity, subject to the completion of certain milestones and conditions related to robotaxi development.

As outlined on Slide 14, this brings total available liquidity and expected capital in 2026 of nearly $8 billion. Additionally, we're very excited to partner with the U.S. Department of Energy to grow our U.S. manufacturing footprint. The up to $4.5 billion DOE loan, which consists of approximately $4 billion of principal and approximately $500 million of capitalized interest provides low-cost financing for our 300,000 unit capacity greenfield expansion in Georgia, bringing Rivian to meaningful scale.

We expect the 515,000 total units of capacity between our Illinois and Georgia plants will provide Rivian a path to free cash flow positive once fully ramped. We expect to draw on the loan by early 2027, subject to certain conditions. Two weeks ago, our normal factory sustained damage from a tornado. I'm proud of the way our teams have rallied together to get production back up and running while we repair the damages. Despite the weather impact, our 2026 guidance remains unchanged.

We continue to expect full year deliveries of between 62,000 and 67,000 total vehicles across R1, R2 and our commercial vans. We also continue to expect to deliver approximately 9,000 to 11,000 vehicles in Q2 as we expect the ramp of R2 deliveries will be back half weighted. While we continue to believe our gross profit will increase year-over-year, we expect the complexity of a new vehicle launch will negatively impact our automotive gross profit in the second and third quarters before becoming a benefit for our overall operations in the fourth quarter as we ramp production and deliveries.

As a reminder, we believe this is a transition year for the Automotive segment's path towards long-term profitability as we scale R2. For 2026, we continue to expect an adjusted EBITDA loss of between $2.1 billion to $1.8 billion. While economic and geopolitical conditions, including supply chain and international conflicts pose risks, we remain steadfast in our plans to invest behind key growth drivers. We continue to progress our autonomy road map and the expansion of our sales and service footprint as we scale with R2. We believe these strategic investments will deliver long-term value to our shareholders.

Finally, for 2026, we are maintaining our capital expenditure guidance of $1.95 billion to $2.05 billion. Our CapEx spend primarily relates to finalizing construction and tooling for R2 in Normal, the continued build-out of our sales, service and charging infrastructure and kicking off construction of our greenfield plant in Georgia.

In closing, I'd like to congratulate our teams again for the successful start of saleable R2 production and the strong execution in the first quarter. We continue to believe that R2 and our technology road map will be truly transformative for the growth and profitability of our business.

I'd like to turn the call back over to the operator to open the line for Q&A.

[Operator Instructions] Our first question comes from Shreyas Patil from Wolfe Research.

Maybe first, just picking up on a comment that you made earlier, Claire. If you could help give us some more color on some of the actions you're taking to mitigate the increase in commodity costs and some of the metals prices that we've seen increase recently? And what's been the magnitude of increase, if you could help frame that?

Well, thanks, Shreyas, for the question. Yes, we're spending a lot of time, of course, focused on all the changes that are happening from a supply chain point of view. And in terms of raw materials and some of the cost of metals, specifically aluminum, this has been a big focus for us. Fortunately, we've -- our sourcing team has been sort of -- we've grown our sourcing team and evolved our sourcing team over the last handful of years where supply chain continues to be an area where there's a lot of unknowns, there's a lot of variability and a lot of a need for us to be very hands-on and very proactive. And so we've been proactive in both our relationship with existing suppliers, but also in making sure we have, particularly in some of these key commodities, alternative sources of supply.

Okay. Great. And then maybe, Claire, just to clarify, I think you've made a comment about how -- when Normal and the Georgia facility are fully ramped, you'd be getting to free cash flow positive. I just want to make sure if I understood that correctly. And if that's the case, that could be quite a while from now. So maybe just help us understand sort of the trajectory of CapEx maybe near term, but then also as we kind of think ahead and you start to kind of put more in the ground at Georgia?

Sure. Shreyas, the comment that I made on the Rivian's ramp up, its Normal facility plus Georgia facility is what takes Rivian to free cash flow positive in the future. And as we talked a little bit about in our prepared remarks, importantly, we have the $4.5 billion of capital from the Department of Energy loan, which provides up to 80% loan-to-value against the build-out of our future Georgia facility. So while we certainly will see an anticipated increase in our capital expenditures as we approach the start of production in Georgia, we do have significant offsets from a capital road map. And the $4.5 billion is just one component of the full $13.6 billion of total liquidity and expected capital through both the cash that we have on hand on our balance sheet, the added availability of our ABL facility, and then the expected capital from our partners with both Volkswagen and Uber that we expect to receive over the coming years as well.

Okay. Great. And maybe just one quick clarification. The DOE loan, has there been any change to that? I think the original amount was $6.6 billion that was available. Just curious if there's any change there.

Yes. So included within our financial release today, we provided an update on the Department of Energy loans. So we'll now have $4.5 billion of loan capacity that will go towards the build-out of our first phase of capacity expansion in Georgia. We've also increased the capacity of the Georgia site, the initial capacity from 200,000 units to 300,000 units. And so the comment that I made in my prepared remarks is really the importance of the funding road map, especially as we think about the Georgia site specifically taking Rivian to meaningful scale in the future.

Our next question is from Joe Spak from UBS.

Claire, just to maybe pick up on the DOE loan part and to clarify, like obviously, this first phase is an increase from that 200,000 to 300,000 units. But and again, admittedly just being able to skim the document, it does seem like maybe the total project scope is now capped at 300,000 units versus -- I know before it was supposed to be 200,000 units -- Phase 1, 200,000 units, Phase 2 for 400,000 units. So I want to make sure I understand that. And then if you still see an opportunity over time to grow further in Georgia.

So the strategic decision that we took was to increase the initial phase of production capacity to the 300,000 units. On our Georgia site, the full initial capacity will be put on the upper pad at the site. So we have the lower pad, which is still going to be entirely untouched greenfield for future expansion.

Okay. But -- and that might be funded more organically in the future, not necessarily with loan or it's TBD, I guess. But the loan is really only up for the 300,000 units, correct?

So the loan is for the initial phase. The important piece is we've increased the loan size associated with the initial phase as we've also scaled the production volume as well.

Okay. And then just I appreciate the comments on input costs and -- but I guess the other thing that I was wondering about was with tariffs, and this is sort of come with a lot of companies thus far. Can you remind us like what you've paid in IEPA roundabout over the past year? And have you filed for any reimbursement? And was anything booked in the quarter related to any potential reimbursements?

We did not book anything this quarter associated with IEPA tariffs, but we do believe that the recovery of those IEPA tariffs is possible in the future. And I contextualize the sizing to be in the tens of millions of dollars of future benefit.

Okay. But that -- and is that considered at all in your reiterated outlook or that would be upside? I guess it's not significant.

I would characterize it as considered within our current outlook.

Okay. And then just lastly, like I know you talked with Uber, you talked about pulling forward raising R&D in '27. Does any of that work start to seep into '26? And is there a change to the R&D outlook for this year? Or it's really more of a '27 factor?

You'll see the pace of acceleration increase in terms of the spend towards autonomy in '27, but we'll certainly see acceleration throughout the course of this year as well. If you look at Q1, our cash R&D expense increased about 22% this year for that quarter. You could directionally think about that as being more of a year-over-year type run rate as we look out over the remainder of the year.

Our next question is from Itay Michaeli from TD Cowen.

Just first, going back to the Georgia capacity optimization. Curious if it has any impact on your previous long-term financial targets of 25% gross margin. And maybe on that as well, if you can maybe share your initial kind of takeaways on kind of R2 demand generation since you kind of launched the trends.

Well, thanks, Itay. Yes. And I think, obviously, the decision to increase the capacity of the first phase in Georgia coincides with -- I should say, it reflects the level of confidence in our products and our business. I think most importantly, we've just started production on R2 out of our existing Normal, Illinois facility, and we've had early media events and early customer events and the level of enthusiasm for the product has just been outstanding.

So everything from the packaging of the vehicle to the way that it drives to the integration of technology, the overall response has been overwhelmingly positive. And so that bodes extremely well for the ramp-up happening over the course of this year and into next year, but it also sets up a wonderful foundation for us as we think about further capacity on this platform, both for R2 as well as R3 and variants of those vehicles out of the Georgia facility.

Terrific. And then maybe as a follow-up on the Uber announcement. I'm curious whether the robotaxis themselves that will go into the Uber network will have the kind of exact same hardware set as the personal vehicles. And I ask because if you're going to launch in 2028 in complex domains like San Francisco and Miami, would that not also imply a pretty wide ODD for the personal vehicles if they're both operating on the same hardware?

Yes. We talked about this during our Autonomy Day late last year. But I think it's important to recognize there's going to be a whole series of steps we make in terms of progressing towards Level 4. And so in that series of steps, the first later this year on our consumer vehicles is launching our point-to-point capability. And so that's the ability for the vehicle to drive entirely on its own to an address.

And I just this week had -- we do lots of regular rides internally, and I had a great ride with James and the team. And it's so exciting to see how much it's progressed and our technology has progressed even since our Autonomy Day late last year. And so we're very encouraged by this. But that first step of making point-to-point available to customers is going to be a really important step for our consumer vehicles.

As we continue to go into 2027, we'll be allowing in specific areas, eyes off. And so it's hands-off, eyes off, that's a Level 3 capability. And then as we go into 2028, as you said, that's when we'll have our first deployments of a Level 4 capability in a robotaxi. And in the robotaxi variant, there will be some additional sensing on the vehicle. So it will be different than the pure consumer vehicle.

But we are planning to have a personal version of Level 4 as well. And we've talked about that quite a bit. We think the market for a vehicle that you own being able to completely drive itself, do things like drop you at the airport, go to the grocery store, get groceries for you, pick up kids from a sports event. These are really high-value creating activities for the Level 4 capability, and we see them on both robotaxi applications and on personally owned applications.

Our next question is from Dan Levy from Barclays.

I wanted to first start with R2 and the path to getting to positive gross margin, which I think you said would be by the end of the year. Maybe you could just walk through the gating factors. Does -- even with the raw mats, do you still have the confidence you have the right BOM to achieve this? And what milestones do we need to see to make sure that the production ramp is still on track? What are the sort of most limiting factors that you still have to address on this ramp?

We've talked a lot about the cost structure of a vehicle and a huge component of this is, of course, the bill of materials. And the bill of materials is different than the non-bill of materials COGS is contractual. So these are negotiations that happen across hundreds of suppliers, and very different than when we sourced R1. We went into the R2 sourcing with a lot of momentum and much better supplier leverage. And just the level of confidence in Rivian as a business and the level of excitement around R2 helped us put together a set of suppliers that are both very enthusiastic, but that's demonstrated through attractive commercial terms.

And so as it stands, the bill of materials for R2 is about half that of R1. And there's, of course, things we can't predict like raw material changes and DRAM shortages, but the vast majority of the BOM is very stable, and we have a lot of confidence in being able to achieve that -- the target BOM, which supports the very healthy gross margins we've talked about in the past.

Now with regards to the plant, I'll invite Javier just to comment on some of the progress that's happening in terms of ramping up over the course of the next several months.

Yes. Thank you, RJ. Indeed, as you explained some minutes ago, we had last week, the celebration of the first saleable builds and delivers to customers this week. So very proud of the situation we are achieving now. The industrial process is ready. The people is ready as well. We have been through the right training and build cycles. And I would say plant is prepared, process are defined, and we are very confident in our capability to deliver.

I feel confident as well regarding what is our team in place. We have brought in a group of seasoned leaders that have done launches back in the past, big experience on that area. And we are, on the other hand, managing the supply chain, making sure that the supplier scales with us. We have boots on the ground supporting some key suppliers. And we are doing this with our mindset and supplier relationship of transparency and collaboration. Resilient supply chain, agility and intelligence are key factors for success.

Great. As a follow-up, RJ, I wanted to double-click on the point you gave in the prior question from Itay about getting to L4. You'll have point-to-point at the end of this year, and you'll only have the vehicles with the LiDAR end of this year, beginning of next year. It does seem like there's probably a lot of testing that has to happen between when you get those cars with the LiDAR out to the point where you have a launch. So just help us understand what the testing curve looks like, what you need to do from when you have the cars with the LiDAR to being able to unlock L4 because it does seem like there's a lot of miles that have to be driven on that new vehicle.

I think a really important point to make here is just the way the self-driving system is architected. So this is the platform that we launched actually on our Gen 2 R1 vehicles is designed around an end-to-end approach where we're building really what we call a large driving model, but think of it as a neural net or a foundation model for driving. And that model is being fed with all of our Gen 2 R1 vehicles and of course, our launch R2 vehicles and ultimately, as you said, R2 vehicles that include a LiDAR, but very different than previous architectures around self-driving where there were rules-based and more classically controlled.

As you add more perception and as you add more compute, the capability of the model only grows. You don't lose the previous knowledge embedded in the model. And so I often sort of compare it to imagine if you learn to drive with that vision and then I hand into your pair of glasses. You wouldn't forget your knowledge as a driver, you just suddenly be able to see and perceive things that you may have missed previously so you become a better driver.

And then imagine we could hand you a 10x multiplier to your compute capability effectively makes your brain 10x smarter. Again, you wouldn't forget what you knew before, but suddenly, you start to notice new patterns in more nuanced ways than you had in the past. And so that's very important to recognize is a very fundamental difference in how the model is built today and what we're creating versus the more AV 1.0 stack where they were, as I described, they're very rules-based and very classically controlled.

And so because of that, the data accumulation that's happened already on R1 and that will continue with the growth in our car park with R2, all feeds into our overall LDM into this large driving model. And even as we think about introducing new sensors, things like our LiDAR, this is not as if it's first on the vehicle when it's delivered to customers. We have lots of prototypes today that are running with those. If you're in the Bay Area and happen to be anywhere around Palo Alto, you'll probably see lots of Rivians with a lot of additional sensors and that's part of a ground truth fleet that, again, is feeding into this large driving model to accelerate the speed at which that model is learning. Think of it as a brain, the speed at which we're teaching it to drive.

And the work that will go into ultimately launching a customer-facing version of point-to-point, which today, I was -- as I said, I was in one of our cars driving around both point-to-point earlier this week, it's really exciting. But we want to have when it launches to customers had to be extremely robust. But all that work is accretive to what ultimately will be going into our Level 4 platform.

Our next question is from Andrew Percoco from Morgan Stanley.

Maybe just to start on the commercial side of your business. It looks like Amazon made up almost 50% of your auto revenue in the quarter, so a little bit above historical run rates. Can you just maybe talk to what you're seeing with that relationship and maybe even outside of Amazon, the level of maybe interest you're seeing in the commercial product since you launched that extended range version of the commercial vehicle?

Yes. Our relationship with Amazon continues to be something that we're very proud of. We've spent a lot of time on this program from its initial kickoff quite some time ago in 2019 through its initial launch and now ramping, deploying. That's everything from not only building the vehicles, but on the Amazon side, getting their operations and the infrastructure ready to ingest a lot of EVs. And what we're now seeing is a reflection of all that work, all the cumulative work that's happened to date that's allowing the volumes for our van program within Amazon to grow, as you point out, pretty meaningfully. And we expect that increased demand for vans to continue. And that's super rewarding to see. It's fun to see all the vans on the road, but that's going to continue to ramp up with Amazon.

Now in terms of other customers and other applications, of course, Amazon is by a significant degree, the largest operator. And so they're the ideal lead customer, if you will, but there are lots of other opportunities we've seen. But in the immediate term, our focus remains on Amazon and ramping to support them.

Okay. That makes sense. And then maybe just -- I just want to ask one more question on the DOE revised loan piece here. I understand the movement in Phase 1 and upsizing that. I'm curious why you might not want to use the DOE funding for the eventual Phase 2. Is this something initiated on your end? Or did maybe they approach you in terms of revising that? Just kind of curious the thought process around why not tap that low-cost funding for the eventual Phase 2 whenever that comes about.

Thanks, Andrew. As I mentioned in my prepared remarks, we're really excited to partner with the Department of Energy on Rivian's $4.5 billion loan, which enables thousands of American jobs and helps us establish the U.S.'s strength in technology and manufacturing leadership. The DOE loan is uniquely a very cost-efficient form of capital as we spent a little bit of time walking through Rivian's broader road map. But specifically, the importance of this $4.5 billion is the funding of Rivian's scaling its operation up to 515,000 units of overall capacity and the opportunity with that installed capacity base to be free cash flow positive in the future. We'll continue to be opportunistic as it pertains to our capital road map beyond the components that I had outlined in the existing $13.6 billion of liquidity and total expected capital that we've outlined today.

Our next question is from George Gianarikas from Canaccord.

So I know it's early days, but I was wondering if you could please give us any color on R2 order trends and maybe some color on the conversion ratios relative to previous orders.

George, as you said, it is early days for deliveries, but the signals I'd be looking at are just the reception around the product and how -- whether it's expert journalists, automotive journalists or lifestyle journalists or customers that are getting to experience the vehicle, the overall excitement around what we've been able to put together in terms of content features, packaging, just the overall value proposition is really resonating. And I'm really pleased with having spent a very large amount of time in the car and it's my daily driver. I couldn't be more pleased with the result. The work that the teams did to make something that's truly remarkable. And we had a few journalists say this might be the best vehicle ever made. That's wonderful for the teams to hear, and it's really encouraging for us as we get ready to ramp the vehicle.

And maybe just as a follow-up, I just wanted to confirm that the Gen 3 sensor suite was going to be available later this year on the R2?

That's correct. Yes. So the Gen 3 autonomy hardware suite, which is both our in-house RAP1 platform. And so this is our in-house inference platform, 800 tops per chip. We have 2 of those chips in the vehicle. So it's extremely powerful. It's a big increase, roughly a 4x increase relative to the NVIDIA-based platform. And then the inclusion of LiDAR, as was referenced before, along with some other enhancements across the rest of the perception stack.

Our next question is from Mark Delaney from Goldman Sachs.

Starting on Autonomy, I'm hoping you could provide more details on the monetization of Autonomy+ so far and any data points you can share on that? And what that might mean for growth in the Software and Services business more generally, including if you still think you can grow that segment revenue by about 60% this year?

We're encouraged by what we're seeing in the -- as you noted at the start of having paid Autonomy+, and it's exceeding our own models on this. So we're -- the take rate is higher than what we expected. And that bodes really well for us as we're going to be growing the feature set quite significantly over the course of this year. And the introduction of point-to-point, we think is a major value driver for customers.

To be clear, this is -- as I said, you can put the address for the location you're going to into the car and the car will fully drive you there. And then following that, allowing you to go eyes off in highway conditions and ultimately everywhere, that means you get your time back. And so we're very bullish on the long-term trajectory to monetize our autonomy on the consumer side. And that's for both hands off, eyes on, hands off, eyes off and then ultimately, Level 4 for personal consumption, we see as a really key driver of value in the long term.

Now in terms of what that does for our Software and Services growth, that is going to start to be something we'll see, but it's not something that we're going to be breaking out separately.

And then my other question was on demand for the R1. I'm curious if Rivian has seen any improvement in order rates for R1 maybe in response to the recent increase in gasoline prices and what that might all mean for R1 volumes this year? I think the company had assumed R1 would decline. Is that still your expectation?

We're encouraged by the continued enthusiasm for R1. It continues to be one of the market share leaders in the premium category. And in a number of states, it's the best -- not just one of the best-selling premium electric cars, but one of the best-selling premium SUVs, electric or nonelectric. So that's true in a handful of states, we've talked about that in the past.

I think it's hard to say ultimately what's going to happen around demand with the impact of gas prices going up. Of course, it's a consideration, and we do see that manifest in what people are trading in. We're seeing more trades of gasoline vehicles or vehicles are less efficient than what we're building. And so we do see that on the rise. But I think a lot of folks are wondering how long fuel prices are going to stay high like this.

Our next question is from Andres Sheppard from Cantor Fitzgerald.

Congratulations on the quarter and all the great progress. I think a lot of our questions have been asked. But RJ, I want to go back to a topic I know you're very passionate about, which is Autonomy. And so I guess with R2 beginning customer deliveries over the coming weeks and with the Autonomy+ now having started this month. Just curious on kind of your vision and how you are thinking about that Autonomy customer adoption? What type of Autonomy penetration rate do you expect for your customer-owned vehicles? Do you expect customers will prefer the monthly subscription or the onetime purchase? Just any color here on your overall vision and customer penetration adoption that you might be expecting?

Yes. Andres, we're extremely bullish on the importance of Autonomy for customers over the next, call it, 5 years, and the rate at which we see customers adopting and selecting Autonomy+. And then ultimately, as the feature set grows and the capability grows, that adoption rate growing with it. But I think that there's an even bigger question just from a society point of view, we've been on a journey of -- when we think about Autonomy where Level 2 is where your eyes are still on the road, but you're still responsible for driving the vehicle. That's like a small appetizer for what you can actually achieve when you get to higher levels of Autonomy where you can take your eyes off the road and truly get your time back and get your time back without the car, dinging you to say, hey, look back at the road or pay attention or put your hands on the wheel.

And so as that starts to occur and as people start to experience what it's like to truly have your time back, so take a 40-minute commute and the idea of getting those 40 minutes back in both directions. We think it's going to be a very sticky experience. And it's going to be something that once you experience it, even if it's indirect, let's say, in a friend's car, it's going to become a very important purchase criteria.

And the reason I call this out is we really believe over the next 5 years, the rate of progress of what we're going to achieve with Autonomy will look very, very different, and I'm talking here at an industry level versus what we've achieved over the last 5 years, means that the topology of customer expectations and therefore the way that the vehicle purchases are made and the criteria that are being used to make those vehicles is going to look very, very different in 2030, 2031 than it does today, where Autonomy will be a very critical criteria where customers are willing to pay for it because they want their time back. They want to not have to be paying attention. They want to be able to be on their phone, reading a book, taking a nap, truly getting time back while you're in the car.

And so as a result, as you've heard a few times throughout this call, this is an enormous focus area for us as a business. We're very much deploying a lot of our R&D dollars towards this category, and we've made long-term investments in the hardware and the vehicles to support that.

Got it. That's super helpful. I really appreciate all that color. Maybe just as a quick follow-up, one for Claire. So just regarding your delivery guidance for this year, which is unchanged. I know in the past you've given us some cadence for deliveries in Q2. I think you reaffirmed that on the call just now. Can you just remind us what kind of unit mix we should be expecting for the year across R2, R1 and EDVs? I think in the past, you might have mentioned R1 and EDV is relatively flat. So the delta should be the R2. But I guess for the R2, the $45,000 price range, right, that's on track for next year. Just curious if you can maybe remind us how we should think about unit mix for the rest of this year.

Sure. As you reiterated, as you think about the composition of the 62,000 to 67,000 deliveries, we anticipate R1 combined with the commercial vans to be roughly flat relative to our 2025 delivery results and then the remainder being comprised of the introduction and ramp of R2, which as implied by the 9,000 to 11,000 of Q2 deliveries suggests more of a back half weighted ramp associated with R2, which is implied within our outlook and guidance.

Our next question is from Edison Yu from Deutsche Bank.

I wanted to come back on robotaxi. Are there any sort of KPIs that you're sort of tracking or that you need to hit for some of the milestones with Uber? And I think in the past the industry has kind of turned to disengagements or miles between intervention. Any flavor on that would be great.

On the path to deploying in 2028, there are a number of milestones and some of those tied to the investment unlocks with Uber. The first of those is later this year we'll be deploying vehicles in both San Francisco and Miami with a safety driver. So the vehicles will be running, but with the benefit of a safety driver in the vehicle with them. And there's a handful of additional milestones ramping up to ultimately having the vehicles operate fully on their own as part of a service in 2028. But as we get closer and closer to that date, there will be -- there'll be lots of proof points, if you will, of the progress that's being made that will manifest on the road. You'll actually see them not only being tested, but you'll see them as part of some of these deployed fleets.

Understood. And just a kind of a separate question on more Autonomy more broadly. I think there were some reports that you guys were looking to potentially license some tech to other OEMs. Just wondering anything you can say about that? Is that something that we could potentially expect this year? Anything that would be great.

I think there's 2 broad categories of technology we think that as I referenced earlier that will be very, very important for growing or maintaining market share in the next several years. And so the first of those is shifting away from a domain-based network architecture where you have a very large number of supplier sourced ECUs that are -- think of them as little islands of code on little small computers where you might have, depending on the car anywhere from 50 to 150 of those little ECUs, those little computers that are in aggregate, providing the software for the vehicle.

But that architecture is incredibly hard to do updates on. It's very, very difficult to do cross-platform or cross-domain integrations. It makes the idea of integrating in a deep way AI into the vehicle and the vehicle experience very difficult borderline impossible to do it well. And so our view is every vehicle on the road will need to shift to a much more centralized compute where you have more of a zonal architecture. So essentially, think of it as a large consolidation of all those little computers into a single or a very small number of large computers that runs a common operating system and for which the code base that's running on the vehicle can be very easily updated without coordinating among, let's say, 50 to 150 suppliers.

And so that architecture I just described is, of course, what's in the Rivian vehicle today. It's also the basis of our relationship that we've forged with Volkswagen Group to deploy that technology. And the first application of that technology being deployed outside of Rivian as part of our partnership with Volkswagen Group is going to be in the ID1, which is -- it's an EV that will be launched in Europe with a price point of just over $20,000. And I can't wait for people to buy this car. I'm sure lots of people to buy it and take it apart and tear it down. And I'm certain they'll be blown away with the elegance of how we've executed the ECU -- or I should say, the network architecture and the compute stack topology.

And so that's one technology category or one area that we think has a lot of opportunity to be deployed in other manufacturers. And of course, the proof point that we're successfully deploying this into a very large established manufacturer within Volkswagen Group across multiple brands, across multiple price points, different form factors, different geographies is the proof point or the existence proof that the technology is scalable and that we're capable of supporting these types of complex deployments.

With that said, the other large category of technology that we see opportunities to have licensing deals is in the autonomy realm. And here, it's in a similar way, it's not just hardware and it's not just software, it's the 2 of them together. So it's the combination of our compute platform that we've developed, the RAP1 in-house inference platform I talked about and the associated computing platform we've designed around that, along with our perception platforms of the cameras, radar, LiDAR that exists. And then very importantly, the large driving model, this foundation model, this neural map that we've created that defines what driving or how to drive a vehicle.

And that's -- as hopefully, I made it clear before, is a much more flexible architecture to deploy into different vehicle embodiments. And so we're doing that already within Rivian. We'll be deploying that across R1T, R1S, R2, ultimately our commercial vans, robotaxi applications, but we do see this as a very scalable technology that can be deployed in many ways.

Our next question comes from Alex Perry from Bank of America.

Just one, I guess, a little bit further on the robotaxi strategy. So you have the Uber deal announcement. I guess is the plan to pursue a partnership model for now? Does the deal with Uber have sort of any exclusivity? And how else will you sort of look to tap into this important market?

When you think about the robotaxi space as a business, and so putting aside the technology for a moment, and I should say, and putting aside and recognizing that the technology for Level 4 in a robotaxi or in a personally owned vehicle is the same, meaning a personally owned Level 4 vehicle can't drive -- or should not drive worse than a robotaxi Level 4 vehicle. They're both very capable of managing the same levels of complexity and the same types of driving situations. So it's the same tech stack.

But when you think of it through the lens of a business model, the benefit of working to deploy this first with Uber is you have very large density of choice. And so if you're deploying entirely on your own, you have to build enough vehicles to have in the fleet or in the car park such that if you're a user, when you say I want to have a vehicle available, it's immediately available or it's available in a matter of minutes.

And so the scale of Uber's platform and the success they've had in creating a really healthy marketplace really makes them an ideal partner for us as we think about launching this technology in an R2 to deploy into providing robotaxi services. And as you've heard me say a couple of times, and we've talked about this in the past, that technology is going to also underpin a consumer personally owned variant as well. And I think we have to recognize that there's going to be lots of innovation around business model that start to emerge as we have Level 4.

And so if you think of it in a very simple sense, the 2 bookends in terms of business model are pure ownership or the vehicles dedicated entirely to your household. And the other end of the spectrum is purely mobility as a service, where you don't own the vehicle, you're not using the same vehicle every time, but you ask for a vehicle on a purely variable basis and it shows up. There will be things that emerge in the middle and not to be exhaustive here in the types of things, but you can imagine different forms of sharing vehicles amongst families or within neighborhoods or within apartment buildings. But there's going to be a very exciting time of innovation in terms of how we think about consuming mobility or consuming transportation.

And with all that said, just recognizing the trillions of miles that are driven today, the vast majority of those are driven in personally owned vehicles. And so robotaxi represents a portion of those, but we think there will be lots of new models that start to make up the topology of those trillions of miles that are driven.

Our final question comes from James Picariello from BNP Paribas.

So I want to ask about the Uber partnership. Can you share any color on the milestones that are associated with the 4 tranches of funding, right, regarding the $950 million in additional additive liquidity. And it does appear right that one of the tranches is already expected to hit this year, $250 million?

Yes. As RJ had just mentioned, there are a handful of milestones and the milestone that we more specifically expect to be in position to unlock the initial $250 million this year will be the operation of some Rivian vehicles in San Francisco and Miami with safety drivers later this year. And then as you think about the subsequent years, you can think about the ongoing trajectory towards full deployment in a couple of cities in 2028 and then 25 cities by 2031, that would fully unlock the remaining $700 million of capital from Uber.

Excellent. That's great color. And then just to maybe level-set expectations on automotive gross margins for the second and third quarters. I mean, this quarter was yet again another strong showcasing of the company's momentum toward positive auto gross profit, right? And this is the last quarter before the R2. Like is there anything you could share for these next 2 quarters regarding the temporary order-of-magnitude impact we can expect to auto profitability?

Sure. As we think about the subsequent quarters of Q2 and Q3, we'll see the introduction and turn on of both, all of the depreciation expense, the new manufacturing team that is established that will be producing the vehicles. But as they're in the process of ramping up the first shift of operation, we'll see some of the complexity associated with lower volumes on the new R2 line. And so as a result of those attributes, we do anticipate seeing an impact to our automotive gross profit over Q2 and Q3 before we start to see the overall benefits of the ramp, not just on the R2's unit economic profile, but also importantly, the fixed cost leverage that we'll see across the R1 program and EDV program overall.

So in total, we still anticipate that we'll exit 2026 with a trajectory of positive automotive gross profit with that being both R2 as well as total Rivian Automotive gross profit being positive, which is important for us as we go into '27 and really fully ramp up the R2 capacity in Normal.

This concludes the Q&A section of the call. I would now like to turn the call back to RJ Scaringe for closing remarks.

Thanks, everybody, for joining us today. Hopefully, you can tell we're really looking forward to getting R2s into customer hands. We're very pleased and excited with the product that we've developed and proud of the team for all the great work that went into creating such a special vehicle. Along with that, we are very much focused on the development of our Autonomy platform. And with that, we'll be starting to see some of the fruit of that significant effort, as I said, first, with our point-to-point capabilities later this year and then adding more functionality, more capabilities over the course of 2027 and '28.

And again, thank you everybody for joining this call. We're excited for all of you to hopefully experience an R2 and see them on the roads here very soon.

This concludes today's call. Thank you for joining us.

All right. Up next, we are pleased to be joined by Claire McDonough, CFO of Rivian; James Philbin, VP of AI and Autonomy at Rivian. Before -- so thank you for joining. Before we get started, just important disclosure here. So for important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. All right.

With that out of the way, Claire, maybe I'll kick it off with you. Would love for you just to set the stage for us. A lot's changed with Rivian over the last few years, but it's also a very pivotal year with the R2 launch. So what's changed? What's working better? And what do you think matters most for investors over the next 12 months?

Sure. As you look back at Rivian's history, our intention was to bring the R1 program to market that was really going to be Rivian's handshake with the world. That was going to showcase the performance, utility, capabilities that could be captured in an electric adventure vehicle. And I think the R1 has done that for us in spades. So the R1S, for example, is the best-selling electric SUV over $70,000 across North America. And in the state of California, it's actually the best-selling premium SUV sort of combustion engine or EV as well. So we need more states to sort of take a cue from California in that regard as we think about the adoption rates for the future as well.

But as we think about what's coming next with R2, if we can just harness even a fraction of the market share and performance we were able to achieve with R1 in a much larger addressable market with R2. R2 will be a home run beyond our wildest dreams as well. So it showcases the intentionality of the product road map.

And then beyond the product road map itself, R2 is also -- you can think about it as a vessel for all of our advanced technology. So it's not just a vehicle program, but it's our next generation of electrical architecture, software. And then I think one area of the business that has probably made the furthest advancements or changes since we started with our launch products is centered around autonomy with full in-house hardware and software stack, which James is going to tell us a little bit more about today as well. So that's critically important as we think about R2, which represents the scale, growth and profitability potential for the business. But also, I think one of the other elements that has evolved over Rivian's history is seeing Rivian as a technology provider for the industry at large.

And that is illustrated through the $5.8 billion joint venture that we did with Volkswagen Group and now the opportunity for us to potentially license autonomous hardware and software capabilities in addition to the electrical architecture and software to other OEMs across the industry as they see the need for transformation and the opportunity set in a world where in the not-too-distant future, if you don't have a software and AI-defined vehicle with advanced autonomy, you're not going to be able to generate the share position that you would want to have in the industry as well.

Yes, that makes sense. And we agree with you that it feels like autonomy is really the driving force behind EV adoption going forward. And it become synonymous with buying an EV is what's the autonomous driving platform that you have layered on top of the vehicle. So James, maybe I'd love to pass it to you to discuss how does Rivian differ? What's the strategy been when you think about building out your autonomy platform? It feels like every OEM has a slightly different strategy in terms of how they're approaching it. So I'd love to hear how Rivian differs, what gives you the confidence on kind of the road map going forward on your autonomous driving offerings?

Yes. So I think the first thing to point out is that Rivian is really a developer of the technology. We're not just an integrator, right? So I think most OEMs when they're talking about their L2+ or their ADAS solutions, what they're really saying is that they're taking a solution from a Tier 2 or a Tier 1 and integrating into their vehicles. I think we're very different in that we really have built this sort of integrated end-to-end approach all the way from the sensors, all the way through to the compute and then the software that runs there through to the experience that the customer has.

So I think that gives us a tremendous sort of nimbleness and velocity that other OEMs don't have. With the launch of Gen 2, we spent a lot of time building the right architecture, the right foundation really around the sort of data flywheel approach, right? So every vehicle that's on the road today from the Gen 2 R1s actually has a very powerful collection -- data collection mechanism in it that can trigger, can pull that data, can learn from customer driving. And so all of that feeds back into the system.

So we launched Gen 2 in 2024. Early last year, we launched our first hands-free product. And because of that flywheel, we were able to massively expand that with the launch of UHF, which was the end of last year. So that's actually the in my opinion, the best hands-free system on the market today, can cover more than 3.5 million miles of roads in the U.S. And so that's just the beginning, right? So we're on this amazing trajectory that this year, we're planning to get our first customer in a point-to-point system. We demoed that actually at AI Day and really building towards those much higher levels of autonomy.

Yes, that's great. And maybe just build upon that a little bit on the road map for eyes-on, hands-off point-to-point because I think that's a pretty big step function improvement in the technology. And then eventually, you've talked about getting to true eyes-off hands-off point-to-point. What needs to go right to get there? And what does that scaling look like? Is that a start in a very geofenced area like you did with the universal hands-free highway driving where it's on a select number of miles and then you eventually expanded it? Or would you wait to roll it out until it's available to the broader universe of customers? And maybe a secondary question to that is like how -- and maybe this is more for Claire, but how linked is it to the success of R2 and the data collection from R2? Or do you have enough, I guess, with the R1 vehicles on road from a data standpoint to get to that eyes-off level?

Yes. So I think the way I think about it is that we have in the fleet that's deployed today, a very large funnel. And in fact, it's not like we're recording and sending back all of that data, right? There's -- most of the data is not interesting. So if you just sat in the middle of the road, there's not much going on. We don't need to collect more of that data. So as we've deployed this system and as we've deployed UHF last year, we've also been improving like the triggering mechanism.

So how do we find those interesting use cases, those interesting tail events so that we -- the system learns as quickly as it can. So that's, I think, a big part of it. R2 ramp will definitely increase the volume. And the thing to remember about our vehicles is that they're actually -- we have 11 cameras, 5 radars, essentially one of the best sensor sets of any vehicle on sale in North America. So the data we get back is actually much more valuable than just a camera and data set.

And then with the LiDAR launch later this year, we'll actually have an even more powerful data set that's -- we think of it actually like a ground truth fleet. So we essentially have the biggest ground truth fleet in the world essentially. And all of that will feed back in, make the system much, much better, much, much faster.

That makes sense. And I think there's a debate in the market as well around simulation versus actually real-world data capture. And obviously, a lot of other competitors out there promoting simulation as a way to really expedite that eyes-on or even eyes-off, hands-off driving for some of the maybe more incumbent OEMs that haven't taken this in-house. How do you view that? How much do you leverage simulation? And are you a believer that this needs to be done on the road and like in the real world to get to a point where you solve that 0.1% edge case that you need to ultimately solve to make sure it's safe? Or is the sim good enough to get there?

Yes. So there's sort of two types of simulations. So there's what I would call open-loop simulation where you actually take a real scenario and you replay it through the system. So we already do a lot of that today. In fact, we resimulate millions of miles even for releases today before any customer sees them. Then there's another type of simulation, almost more like a video game style where you're sort of recreating camera imagery. That, we think, is important, but much less useful, especially for us because we have this incredible diversity of data from all across the country and all these different weather conditions.

Just to give you an example, for UHF that we launched last year, although we haven't done much validation specific work around snowy conditions, because of all that data that we collected since the launch of Gen 2, the system was actually very capable in those conditions. I had multiple people texting me saying, wow, this is the best system in snow. And so it's like that emergent property that we got because we had that incredible diversity. We didn't have to go specifically, oh, collect snowstorm data, collect data here. That already is coming to us.

That's great. And maybe R2 sounds like it's the marquee platform that's going to help you scale to be a scaled player and the R1 demonstrated the technology chops that you guys have. And I think that's very clear. As we think about the R2, you're launching it, I guess, in a week from today. You're starting the configuration process, start to build an order book. What does success look like to you? I know you've given a guidance for the amount of vehicles that you want to deliver this year. But outside of that, is it margin? Is it production time line? Like how should investors track over the course of the next few months as you start to build that order book, whether or not you're ramping or you hit your kind of successful milestones on R2?

Sure. As we take a step back and look at what success looks like, first and foremost, it's expanding the addressable market of Rivian consumers. So tapping into and not just prior EV owners, but also creating new choice where a consumer today may be on the fence in their first EV purchase decision and the opportunity for Rivian to take the 90% of consumers today that are still purchasing a combustion engine vehicle and get them behind the wheel of their very first EV. So that, I think, is critically important as we think about what success of R2 looks like as well.

The other elements of R2 are the operational efficiencies that we deploy as we execute our production ramp and path to profitability on the product. So we believe R2 has a path to exit 2026 with positive gross profit margins and Rivian to be positive on an automotive gross profit basis as well as we exit '26. And a lot of that is enabled through the fixed cost leverage that R2 brings to our existing Normal, Illinois plant, where last year, we delivered just over 42,000 units. The Normal plant once fully ramped with R2 is capable of producing and delivering north of about 215,000 units in total.

And so that enables not just a path to significant gross profit contribution from the R2 product, but also how it helps enable a path to expanded profitability in R1 and the commercial van as well. And we spoke a little bit about it as R2 as being the conduit for many of our next-generation advanced technologies. We also see their success looks like high levels of take rate in our Autonomy+ offering, which is our new paid autonomous offering for consumers, high utilization and engagement with our AI assistant, seeing the consumers that are really taking advantage of a lot of the core functions and features that make their R2 and the Rivian truly unique as well.

Yes. And I want to double-click on the gross profit improvement, the path to profitability with R2. Totally makes sense on the fixed cost leverage, operating leverage side. But talk a little bit more about bill of material. There's been a lot of conversation, especially at this conference about memory pricing and costs and bottlenecks. You got lithium pricing, copper pricing. How do you balance that? Are you comfortable with where you're at in terms of achieving the 50% reduction in bill of material that you guys have talked about historically? Like how has that changed over the last 6 to 9 months as we've seen the commodity inflation really accelerate?

So while we've certainly seen commodity inflation for some of the core raw material inputs in R2 increase over the last year, we also have always planned for a future where some of these elements were intact. So we weren't just building towards the establishment of the at-market pricing of where aluminum may have been or where lithium prices may have been. We're working towards a product set that has durability through different raw material cycles as well.

And so that certainly went into and contributed to the design and development and the enablement for us to create a product that could be roughly half of the material cost of an R1, but certainly is by no stretch half of the vehicle. So one of the early reviewers of the product said, if you had blindfolded him, put him behind the steering wheel, so to speak, of an R2, he actually probably couldn't have told you the difference. He would have told you he was driving an R1, not an R2. And that just exemplifies the core foundational technologies that enable a path through vertical integration to structural cost advantages that allow us a path to provide such a compelling value proposition to the end consumer as well. So that's certainly central to a lot of the trade-offs that we made that enabled us a path to that cost structure.

Yes. And on the fixed cost side, with ramping at Normal, you should get some benefit there. Talk a little bit about your plans with Georgia, right? I think there's potentially a risk that if you ramp too quickly in Georgia, then you're at the same place from like a fixed cost absorption standpoint that you are today with R1 just given the amount of capacity you have. So just talk a little bit about how you're planning on ramping and expanding Georgia and making sure that matches the demand profile that you're seeing with R2 and eventually R3 as well.

Sure. It's important to note that the Georgia plant is going to be a plant that will house production of our midsize platform. So that reflects R2, the R3 product. You can imagine additional variants of those vehicles that will be situated on the same platform as well. And again, that reinforces just the strength of the technologies that sit at the foundation of our product roadmap as well that we can then leverage to meet lots of different consumer choices, lifestyles to a form factor that truly meets your specific needs. So there's certainly significant runway as we think about the size of the total addressable market that the midsize SUV and crossovers address. Georgia will also be a key launch point for us to begin exporting products to Europe and building brands. So it's not just the North American market share and potential. It's really how do we take what's designed as a global platform and take full advantage of that as well over the longer term.

Yes. And on the demand side, maybe I'll pull James back in for a second. We talked about how autonomy feels like it's a big driver for EV demand going forward. We hear that there's a lot of people who own EV today that maybe one of your peers is EVs and that they want to switch, but they want to wait until they have the equivalent of an FSD on the vehicle. How much are you hearing that? And does that make you or do you wish to maybe accelerate your path to launching some of those programs or some of those offerings so that you can compete with some of the other players in the market and make sure that you are really ramping into the capacity and the R2 demand is really going where you want it to go?

Yes, absolutely. I think you see that from consumers more and more as these features get better. And actually -- as well as consumers get more familiar with like robotaxi efforts, they require more of their own vehicle. And so I think we definitely see that in the market and are working very quickly to improve the system to get us to that point. So as I said, with the launch of UHF last year, best-in-class hands-free system.

This year, we're really shooting for that point-to-point early system that will start to get customers in customers' hands end of this year. We actually demoed that at the AI Day. So we're making very rapid progress. A lot of that is powered by this large driving model that is essentially learned from all of the human supervised driving data that we've collected from our fleets. So it's very capable, very powerful. It has this amazing generalization that you can take it to new areas and it kind of just works. So a lot of the work this year will be sort of finessing that system, putting the right guardrails around it so that we can get that product out.

So I think consumers should feel confident that the Autonomy+ today is just the start. We're going to be improving that system to be best-in-class within that package. And so -- yes, and then with R2, with the new chip and the LiDAR, then we can take it to that next level. And I do think that there's probably -- there's a step change coming with -- once we get to sort of true eyes-off, that's something that would really give you your time back on a long commute. And I think there, you'll see consumer preferences really start to shift where that now becomes not just like a nice additional feature, but a must-have. You can imagine people buying not one R2, but two R2s, right, because both members of the household need to have that feature. So I think, yes, it's a very exciting future, I think.

That's great. And when we come back to the demand side, I think you talked about R1 and the commercial vehicle demand profile you're expecting to be roughly stable or flat year-over-year. Talk a little bit about -- you even mentioned that the influencer said that when they sat in the R2, it felt like it was the R1. So how do you balance that from like a cannibalization standpoint to ensure that demand for R1 is still strong given the fact that R2 is much more accessible for a lot of people at that price point?

So one of the distinct advantages we have is truly a different form factor. So there are many households where they need to have a three-row, seven-passenger SUV to meet their personal lifestyle needs. And for those consumers, they're still going to gravitate towards an R1. And for some of the more price-conscious consumers, we also have the opportunity to sell them a used R1 as well as they think about their specific choice. But there will be many consumers that R2 is certainly the right form factor for their personal needs and use cases. And as we look at the evolution of the R1 product portfolio, R1 will continue to skew more premium and create a pretty big gap relative to the price point and sizing of the R2 over time as well.

Got it. And on the commercial side of the business, the commercial vehicle, talk a little bit about what you're seeing there with Amazon and maybe outside of Amazon, the traction you're getting with some new customers. Amazon has been a big customer for you in that vehicle historically. But what are you seeing from others? And maybe what are some of the gating factors to getting some other corporates or commercial customers to actually adopt an electric vehicle as a last-mile delivery option?

This year, we do expect to see growth of the commercial van. That growth is predominantly driven by Amazon specifically. We're launching new variants for Amazon, including a larger battery pack variant of the van and all-wheel drive variant so that they can use the EDV for more rural routes within their network. And this growth is also a testament to the total cost of ownership that the EDV does provide.

And as we think about the opportunity for other fleet operators and owners, it's the continued education of the TCO benefits of managing, operating a software-defined vehicle. So in the case of Amazon, we're also very well integrated into their own software applications. We have a FleetOS, a fleet management software subscription attached to each vehicle in the fleet and are really proud about the opportunity to electrify the delivery space at scale with the largest customer in the last mile delivery.

I think it will take some time as well for other fleet operators to make the infrastructure investments and begin to more holistically electrify their fleets. But back to the conversation on autonomy, I think we will certainly see that as sort of another catalyst that will significantly advantage the TCO of an autonomous commercial van in the future as well.

Yes. Yes. That will be interesting to see how it eventually when all these vehicles are autonomous and presumably, you're planning on rolling it out to the commercial vehicles at some point as well. But on the -- you mentioned software and services. That seems to be a bright spot where you always exceed expectations in that part of your business. Talk a little bit about -- you've got the JV in there and then you've got your core software services business. Talk about the path to profitability and the drivers there in terms of the growth that you're seeing and that you're expecting this year? And how should we as investors or sell-siders expect or model that out going forward in terms of really driving that path to profitability for Rivian?

So in 2025, our software and services was just about $1.6 billion in total revenue. And as we look at the year ahead, we anticipate the growth to begin to approach about 60%. And that's driven both by the growth of the joint venture, but also importantly, as the car parc continues to grow and expand, we see a compounding of many of the other services within the portfolio. So we talked a little bit about the launch of Autonomy+. So in April, we'll start our paid subscriptions within the software and services component and see that as being a long-term catalyst for margin expansion opportunity for us as well.

The other large components outside of the joint venture are centered around our remote remarketing program. So we're very focused on how do we preserve the residual values of our product and play a very active role in the used market for Rivian. We also have a large component coming from our maintenance, repair, serviceability. And then as we've talked a little bit about our FleetOS offering, our Connect+ offering, which is our advanced infotainment offering. And then beyond that, financing, insurance, so there's so much value that can be created throughout the ownership lifetime of a vehicle outside of just the initial purchase. And so being a direct-to-consumer player like Rivian is a distinct advantage as you think about the opportunity to expand and enhance the lifetime value of each potential vehicle and each potential customer as well.

Yes. And it's very clear there's -- you're working on a lot of things, autonomy, R2, eventually R3, Georgia, ramping Normal. How should investors think about how you fund all of this, right? There's a lot of exciting things that you're doing. But how do you bridge that gap until you really ramp R2 and you start becoming to a point where you're self-funding the business? What does that look like? And what are kind of the sources of funding that you're expecting to tap into?

So we ended 2025 with $6.1 billion of cash. There's another $2.5 billion of payments, $2 billion of which we anticipate receiving in 2026 from Volkswagen Group. And then as we think about the build-out of Georgia, the Department of Energy loan is really a construction type project-based finance instruments that you can imagine will help us fund some of the growth of both the product portfolio and plant and facilities there as well over the future. And then beyond that, we'll remain opportunistic as it pertains to additional sources of capital.

Great. I want to pause real quick and see if there's any questions in the room. All right, it's none. You talked about it a little bit in your opening remarks about international potential expansion. What does that look like? Europe is becoming -- it feels like a very competitive market, especially on EVs. Why would it make sense to go to Europe just given that competitive dynamic? And where do you see yourselves really playing in that market relative to some of the competitors there?

We think there's a distinct advantage that stems from being a true clean sheet operator. So having an in-house electrical architecture, software stack, advanced autonomy, a fantastic product design and a great brand reputation as well that allows Rivian to compete not just in North America, but to compete on a more national stage as well. So we were also encouraged by having a number of international reviewers come test drive the R2 as well and the glowing feedback that they provided for the product and excitement. They're saying -- begging us to get the R2s as quickly as possible into the continent of Europe as well. So excited about the opportunity to really stretch the brand even further, and -- and that design also is not just part of R2, but also as we think about R3 and that being a very compelling form factor being sort of a smaller crossover vehicle for the continent of Europe.

Is it just Europe or would Middle East or Latin America, any other markets that screen attractive given the form factor? Or is it just Europe was like the first international plan for you guys?

There certainly is a lot of interest beyond just Europe, but we also need to be methodical as we think about the building blocks of infrastructure and scaling the product in multiple locations across the world as well.

Got it. And you also mentioned the partnerships that you have with Volkswagen and potentially expanding that or finding other partnerships with other OEMs. Can you talk about what that discussion looks like? What are some of the limitations in terms of like why others haven't come to the table and done a similar deal that Volkswagen has done? Just trying to get a sense for -- that ultimately would be a very attractive way to monetize the technology beyond just scaling up internally with R2 and R3. So just trying to get a sense for what's the time line? What are some of the gating factors to actually making that happen with other partners?

Today, only Rivian and Tesla are the two Western OEMs that have a zonal electrical architecture in the product. And as we think about some of the design capabilities that will enable us to reduce the cost structure of R2, the electrical architecture is really central to that equation overall. As you think about other OEMs, we certainly see a world where the future is fully autonomous and software and AI defined. And so you will have to have those embedded capabilities in your product to compete in the future. And so I do imagine that there certainly will be OEMs that are either exploring things in their own stand-alone way, navigating this with Tier 1 partners and ultimately, we'll seek or explore opportunities potentially with Rivian in the future, given how compelling the electrical architecture platform is.

And if you take a step back and you look at what we're doing for Volkswagen Group, which is the second largest OEM in the world and probably one of the more complex ones given the variety of brand and breadth of form factors and price points within their portfolio. Rivian's ability to service the Volkswagen Group is really our best calling card as we think about other OEMs that may be interested in these capabilities in the future as well.

Got it. And to dovetail off of that, James, maybe for you, the RAP1 chip that you guys announced late last year, I think you also talked about potentially finding buyers of that chip. How do you think about the evolution of that as a product, right? Is RAP1 good enough to get you to all the way to like eyes-off, hands-off and beyond? Or will you have to keep iterating on that? And what does that look like from like a spend perspective going forward in terms of the capital intensity of the business?

Yes. So as you mentioned, on the AI Day that we had last year, we announced the -- our in-house silicon effort. That's actually not something we did on a whim. That's a multiyear long effort to design and build and source that chip. And yes, the goal is for that chip to be in R2 at the end of this year. It will actually, as far as we know, be the most powerful automotive computer in the market, certainly in the U.S. So yes, it's really an incredible chip, incredible capability. It's really been designed for these transformer architectures, which are these modern building blocks of LLMs. And yes, we think that, that has certainly the capability to get all the way through to L3 and then personal Level 4.

Yes. Great. And maybe to just close it out with a very high-level question for you, Claire. What do you think the market is getting wrong about Rivian and your story that you obviously portray today? What do you think we're missing? What do you think the market is missing about your strategy and your story more broadly?

I think one of the elements that sometimes is missed or not fully understood is just the core technology advantages that a player like Rivian has. And that's seen through our ability to recruit phenomenal talent like we see on James's team, as we see on our in-house silicon team to bring together the best minds as we think about the evolution of our technology foundations for the future in a variety of different areas within the business as well.

I also think that AI will be a distinct advantage for players that are in the direct-to-consumer space as well, where through the ownership -- the end-to-end ownership control, there are so many elements in which we can ultimately improve the ownership experience while also reducing cost over the longer term as well. But I think some of those key advantages within the story are sometimes overlooked or not necessarily fully appreciated. And as you think about the complexities to build these capabilities in-house from the ground up to have these proven technologies that are operating and driving value to consumers already today and then the future road map as we look to sort of package those technologies into R2 and the opportunity to scale with the mass market program.

Well, great. Well, I think we're out of time here. I appreciate it, and thanks for everyone for joining.

Thank you.

Thank you.

Good morning, everyone. My name is Avi Steiner, and I am the automotive analyst here at JPMorgan. It's our pleasure to have with us today from Rivian, Chief Financial Officer, Claire McDonough, who has graciously agreed to be peppered in a fireside chat Q&A format this morning.

As always, I'm going to start with some Q&A up here, and then I'm going to open it up to the audience for questions, hopefully, if time allows, and then we'll go from there.

Claire, thank you for joining us today.

Thanks for having me, Avi.

Excellent. Okay. Let's start here. You've done a great job. The company has done a great job of reducing the bill of materials for next-generation R1, both year-over-year and sequentially. Can you walk us through how the company achieved those savings?

Sure. As you look at our cost of goods sold for per unit delivered in the fourth quarter of 2025, we saw a $7,000 reduction versus the fourth quarter of 2024. And we were able to achieve that through a significant reduction in our material cost of the product. Those were achieved by both the transition that we made from our first-generation R1 to our second generation, which improved all of the core technologies in the vehicle itself, bringing us to a lower cost basis in our material costs on R1 program.

We also saw a slight increase in terms of our commercial van deliveries in the fourth quarter, so had some additional favorability from a mix standpoint as well. But the important piece is, as you look at the fourth quarter, on a cash basis, we made over $2,000 a unit, and that sets us up really nicely as we now think about the added volumes that will come into play with the ramp and rollout of R2 that will start first external deliveries of R2 in the second quarter.

We'll build for the most -- the majority of the year on a single shift, and then we'll add a second shift of operation. So as we look towards the fourth quarter of 2026, we'll see a key inflection point in terms of the overall production volumes and with that, the fixed cost leverage that benefits not just the R2 that is coming to market, but also importantly, the underlying unit economics of our R1 program as well as our commercial van.

You touched on a lot there, and hopefully, we will get to address a bunch of those things.

I want to shift to the R2. It's expected to be a game changer for the company. The reviews so far have been pretty positive. If anyone hasn't seen them, I encourage you to go online. That's not me saying that, that's other companies.

But more importantly, recognizing launch is proceeding deliberately, as you kind of just touched on one shift, what are the biggest financial and operational challenges you're preparing for as the company scales production? How is Rivian mitigating or getting in front of any potential supply chain bottlenecks?

First off, thanks for the shout out on the early release of reviews. We couldn't have asked for better reviews out of the February launch where we had about 10 media outlets, drive the R2 product. Some of the favorite quotes that came out of that experience from my standpoint were one reviewer called it the Best EV that he's ever experienced. Others said that if you had not told them and just put him in -- behind the wheel of the driver seat, he would have thought he was driving an R1S, but instead driving an R2 with a much more affordable price point associated with it as well.

Now as we look at the road map for R2s ramp, we're critically focused on the supply chain right now. And in early January, we started our first manufacturing validation builds on our lines in Normal, Illinois. So over the course of the last 1.5 years, we built out an expansion in our Normal facility to accommodate a brand-new body shop and general assembly of the R2 product.

And we've been going through the cycles each and every week to improve the production quality, improve our processes, system, training of the team. So it's -- you can think about it really as a continuous road map as we reach salable production on the product first, and then ultimately get to a point where we're inflecting and adding a second shift of operation as well.

And the strategy that we've deployed to run most of this year with a single shift is focused on getting the quality right and the assurance that we're keeping the supply base working in lockstep with Rivian as they increase their production rates and volumes ahead of the second shift and ultimately in 2027, the addition of a third shift of operation for R2.

So with the first shift and really the R2 deliveries, I think, starting in the second quarter of this year, remind us as to how we should think about the margin evolution here from launch to -- I'll use the word steady state. I don't know if you can kind of quantify that. And is that after 2 shifts, 3 shifts? Just how to think about that whole evolution.

For R2, we have the advantage of starting with a low material cost basis on the product. And we were able to achieve that through the demonstration of the brand strength with the development and sales of R1 program as well as our commercial van offering that allowed us to get to much lower levels of -- I would characterize cost with the supply base as a whole as we work through the negotiation of the R2 program over the course of 2024.

And that enables us to have a very fast path to variable profitability on the product. And we anticipate by the end of 2026 that we'll be in a position where both R2 and our overall automotive gross profit will be positive within the business. As we look at the quarter sequential mix, you'll see Q2 as -- and Q3 as being more of a headwind associated with the early stages of ramping up a brand-new product. But then as we look at Q4 and the exit rate, R2 starts to be a net beneficiary of our underlying unit economics.

Okay. Maybe one more on R2, if I can. So I think it's March 12, you'll correct me if I'm right.

Yes.

On March 12, you plan to reveal the different configurations that consumers can select for the R2. And what I'm really curious about, you've had -- companies had pretty good learnings, if that's the right word, from the R1 to guide it, and this is obviously a lower-priced vehicle aimed at different consumers. What are some of the puts and takes in terms of how management thinks about combinations, price points that it might make available to the consumer within that -- the enormous backlog that was once noted for that vehicle?

One of the key learnings that we're employing with R2 is to focus on simplicity. So to narrow the SKU assortment, ensure that we're focusing in a way that enables the acceleration of our ramp and road map for the future.

And so as we look at the conversion of our preorder base of customers, we're going to invite those customers to say, are you interested, first and foremost, in configuring a launch edition vehicle for Rivian, not saying everyone come and configure the exact variant that they want, but instead say, this is what we're building today and you can queue up to put in your preorder for that launch edition variant out of the gate that will help us bring up the supply chain, reduce complexity with the ramp and hopefully be more successful as a result of the simplification of the build plan.

Okay. Terrific. And then Rivian unveiled the RAP1 chip at its Autonomy Day last year. Walk us through the capabilities that provides and the timing in terms of introduction of Gen 3 Autonomy, which is really the genesis of this question. And when will it be commercially available or included in the R2? And relatedly, if I can ask, any thoughts on whether the different tech that is available at launch versus maybe later in the year impacts -- might impact at all R2 sales?

First off, we've introduced to the world our Rivian Autonomy Processor at our Autonomy and AI Day back in December of 2025. The core advantage that we're deploying with an in-house silicon program is the opportunity for us to have rapid cycles in terms of development, a product that is custom built for the universe of heavy camera and sensor-based AI physical world applications, which allows us to use the vast majority of the TOPS or processing power of this chip relative to other players in market that are trying to solve for many more versus just the use case of autonomy or advanced physical AI and robotics.

And then the other element is cost as well. And so we had the opportunity with the talent that we had within our team to recruit a world-class semiconductor organization to Rivian to build this program out for Rivian. Today, we have production and intent silicon that we're testing and developing with our teams, and those will go first into vehicles at the end of 2026.

And then you'll see us scale up the R2 volumes in 2027 with this new Gen 3 hardware set, which includes the Rivian Autonomy Processor chip, but also includes LiDAR, the first LiDARs that will go into Rivian's as well. And this is 1,600 TOPS. And so as you think about the overall compute and processing power, it's twice as powerful as what we have in our Gen 2 products or the first R2s that will go in market as well.

However, with that said, R2 will be at launch, still one of the most advanced vehicles in terms of its autonomy capabilities, the level of compute that's in the vehicle as well. And for those first orders, they'll have full access to a point-to-point navigation. So any road that has a painted line on it, they can use a point-to-point navigation to have the car effectively drive them hands-free to their destination.

That's great. And there was a lot in there, so I probably should have broken that up into a couple of questions.

I'm going to build off of that, if I can. If you could talk about the company's NAND and DRAM needs, how that might be impacted given everybody is talking about the shortage and obviously, the pull from AI. And obviously, you've had plenty of learnings again to use that word, just from your build in R1 and different shortages. Just talk about how you might be navigating that?

Sure. The fundamental challenges within the memory space was factored into our guidance for 2026. And as we think about some of the core advantages of a player like Rivian, we do have the flexibility because of our vertical integration in terms of the electronic control units that we design in-house. We work directly with contract manufacturers on the production of those boards that enables us to be more agile in an environment like this.

So back in 2021, when we were going through the semiconductor crisis, for example, we were able to redesign some of our ECUs to move away from constrained parts within the industry as a whole and gravitate towards new supplier partners that had availability towards Rivian. So we're -- our teams are working around the clock to ensure we have the memory for our products. We do use a sizable amount of memory given the advanced technologies and autonomous capabilities of the product itself. But we also do have, I would say, a core inherent advantage in the level of vertical integration that we have and deploy.

Okay. As a reminder to the audience, I'm going to open up to questions in a couple.

Let me kind of take a bigger picture view here, if I can. There are a couple of cross currents as we see it for electric vehicles at the moment. On the one hand, the administration has effectively unwound the incentives that were meant to spur adoption of electric vehicles.

And on the other hand, we see as consequences, I guess, with the former of the D3 and other manufacturers have begun to meaningfully pull back from the EV product, right? And so I'm sure you don't necessarily view them as competitors, and I don't say that in a pejorative or negative way. But just thinking about how folks are moving to hybrids, there's been some discussion of maybe Tesla pulling back on some vehicles. Just how you think about your position, Rivian's position and these 2 kind of crosscurrents either benefiting or hampering demand for Rivian's vehicles?

So we very strongly believe that the future is going to be autonomous, software-defined. And if you look at the characteristics of those 2 capability sets, they often come in the package of being a fully electric vehicle as well. And so we believe that the steadfast clarity of our focus in the road map in each of these core technologies that we think will transform consumer buying behaviors in the future will put Rivian in a great position as others are actually taking a step back from their electrification initiatives as a whole.

And we also want to just bring the very best product to market. So we often get asked about some of the dynamics in market with EVs. Our ambition as a company is to create the best vehicle full stop. And whether you're coming to Rivian because you want an electric powertrain, you're coming to Rivian because you love the design of the product, or you love the technology of the product or the performance. We think we're bringing a fantastic product to market. And our intention is to take a large percentage of the customers today that are still buying combustion engine vehicles and convince them to buy their very first EV with R2 as well.

Right. I'm going to dig into that maybe a little bit further from a different angle. So RJ on the last earnings call listed a number of states that the R1S, I think, is the best-selling EV. And if you look at those states, and again, I don't mean this negatively, but California, New York, New Jersey, Oregon, Washington, D.C. and Virginia, I think they all -- that was very quick, but they all adhere or subscribe or a part of CARB, right? And they're more EV focused for lack of a better definition.

And really, the question is, how does the company increase awareness and adoption? Everything you just talked about, about how you view this as transformative technology and you want to reach those ICE consumers or those ICE vehicle owners. How do you increase awareness adoption in other states and really countrywide to get them to make the switch?

We've seen a lot of power in the strength of our Rivian community. And the best influencers for Rivian are our customers. So the customers that are going out at their Little League baseball game, talking to their neighbors, talking to their teammates, friends about the experience that they have driving their Rivian, that's the best marketing that we can have.

Now the other challenge is we need more consumers to test drive our products because the best way to showcase the performance capability and utility of a Rivian is to get behind the wheel. We have a couple of Rivians downstairs. So for those of you in the audience that have never driven a Rivian, please take us up on the offer to do just that because you can listen to -- Avi and I go back and forth about the business and strategy, but I truly believe that getting behind the wheel makes the vertical integration, makes the intentionality of our technology stack and the true differentiation of Rivian all come together in one driving experience and package that is incredibly well thought out as well.

So that's another strategy of how do we continue to have more customers grow their awareness of Rivian, whether they're in market today for a vehicle or whether they're considering their very first EV for their next product, having more of those experiences is critically important for us.

And the other element is the continued growth and build-out of our physical infrastructure. So our service infrastructure, our sales infrastructure, our charging infrastructure that supports many consumers in many additional geographies as well.

Can you size the infrastructure you put in place, service centers, et cetera, on the one hand? And then just finishing off that last question, if I can. Is there ever a role for a traditional marketing campaign in Rivian's future? Or this is just you want to keep going on the word-of-mouth experience path, which has been great.

So today, from a physical infrastructure standpoint, we have just under 100 service center locations across North America. We have 35 dedicated, we call them spaces, which you can think about them as experiential retail hubs for the business. And we'll add another 40 locations over the course of 2026 as we scale up for the launch and growth of the R2 as well.

One of the keys from a physical infrastructure standpoint is the opportunity for us to be very asset efficient. And in the case of service, the majority of our service can be done through mobile service applications. So we have hundreds of mobile service vans that are deployed across each of our 100 service locations and the majority of events can be done in -- at your home. So there's nothing more convenient than a Rivian service van coming to your home or workplace that can do the job for you without you ever having to lift a finger or bring the car in for services as well.

And one of the core advantages of EVs is they require very limited services as well. And I think that, that's an area where many new owners are just completely surprised and delighted by the efficiencies and reduction in their total cost of ownership through service overall.

Okay. I'm going to open it up to the audience for any questions. I ask because it's webcast that you wait for the mic to come by, so you can ask your question into the microphone. Does anyone have any questions in the audience?

What's the status of the build-out in Georgia? As a resident of Atlanta, I drive through the I-20 Corridor, I see what looks to be infrastructure improvements around where a site may potentially land, but what's the status of that built-out?

Well, great to have you as a Georgian and adjacent customer in our community.

For the Georgia site, we focused on a lot of the long lead time items. So those are oriented around bringing power and utilities to the site. We've done a lot of work with the state on building a lot of the transportation infrastructure, logistics infrastructure. So there's a new overpass new rail spur that's going into effect as well, such that when we get to vertical integration in -- sorry, vertical construction in 2026, we can go much faster with the anticipation of the very first vehicle builds happening at the end of 2028 at the site.

One question upfront.

Can you delve a little deeper into the Tesla decision to get out of the X&S and what that means for you as an opportunity? And also was your perspective on the logic of canceling those programs. Those were obviously the benchmark programs for that EV platform or had been for a long time before the lower-cost versions. Is there any read-through that you guys have about your platform and vision going forward for EVs generally?

So fully agree with you on Model S and Model X as being benchmark programs. Those were really the first programs in the EV space that showcase the capability, performance, desirability of an electric product.

As we look at Rivian's opportunity now, there's certainly customers that Rivian can recruit and look to continue the continued sales of our R1S, which is a 7-passenger large SUV and our R1T electric pickup truck as well. I can't opine on Tesla's strategic rationale in their own decision-making.

But similar to Rivian, we used the R1 platform as really our handshake with the world or opportunity to build brand to showcase the on-road and off-road capability performance of what a Rivian can do, and then now have the opportunity with our mass market price point with R2 to significantly grow the addressable market of consumers that can now consider Rivian as part of their purchasing decision.

Okay. Great. I'm going to turn it back to myself and ask a couple more questions. Can you update us on the relationship between Rivian and VW on the one hand, a couple of reports have been out there intermittently. But more importantly, can you also remind us in the audience what the time line and milestones are to the next payments from VW to Rivian and both timing and amounts?

Sure. Rivian maintains a very strong relationship with Volkswagen Group. We were incredibly proud to have delivered multiple vehicles back to Volkswagen at the end of 2025 for the start of winter testing. And that's a testament to the extensibility of the software and electrical architecture that we've rebuilt, and the joint venture is leveraging for future vehicles as well.

And so in that 13-month time frame, we were able to modify our Gen 2 architecture, which is what powers our R1 -- our Gen 2 R1s as well as what will power R2 and build that into multiple different brand programs for the Volkswagen group, which is multiples faster than they would have ever been able to execute against, working with their own supply base and ecosystem as a whole.

So it showcases the speed and the utility of the foundation of the IP that sits at Rivian and now at Volkswagen Group as part of the joint venture as well and as we think about the ongoing deployment of these products working our way through multiple Volkswagen Group launches in the as well.

And as it pertains to the incremental capital from Volkswagen, as part of the $5.8 billion of consideration for the joint venture, there's a $2.5 billion of that, that is still yet to come to Rivian. We anticipate receiving $2 billion of that this year in 2026, $1 billion tied towards winter test -- successful winter testing and $1 billion, which is a nonrecourse loan from Volkswagen Group to Rivian as part of the consideration and then another $0.5 billion that will come in equity in '27.

Perfect. And then on your last call, and I think during the Autonomy Day as well, the company talked about the potential to license its software-defined vehicle technology to other OEMs. And beyond what it is doing with VW, importantly to note, is it possible to size the opportunity or how we should think about it here? Is this something that can reasonably be expected before all the VW benchmarks have been reached, as you just discussed, and those vehicles launch in '27? Just how do we think about that possibility?

Sure. There's multiple avenues for Rivian and the joint venture to execute on being a technology partner for the industry. And so within the joint venture itself, we have housed our electrical architecture and software stack that powers the key functions of the vehicle itself. Outside of the joint venture is all of our initiatives around autonomy.

And so as we think about what the opportunity is for Rivian, as I mentioned, we believe that all OEMs will and the future need to have a zonal electrical architecture, be a software-defined vehicle and have high levels of autonomy to be part of the consideration set for consumers or risk losing a meaningful market share overall.

And so we see there being a huge opportunity for Rivian to play a meaningful part in that transition for many other OEMs as well. And through our large commercial relationships with both Volkswagen Group and Amazon have built and are building a muscle of how do we work collectively with large complex organizations and customers as we look to that future state of being a technology provider to the industry.

Okay. Very much a stay tuned. We are at a debt conference. So I'm going to go maybe one on the balance sheet and cash use, if I can.

Based on the company's full year guidance, it's likely Rivian will burn another $4 billion of cash just from EBITDA and CapEx spend, right? And then we have working capital, which is an outflow. I don't think you've sized it, but it's -- you're growing this year, it makes sense it will be something not small. And then you have the incoming dollars from VW. You're in a net cash position today. But just walk us through; a, if my math is right; and b, how you think about your liquidity evolving as you head into '27?

We ended 2025 with $6.1 billion of cash. We'll receive another $2 billion this year from Volkswagen Group. Your math is correct around our guidance for adjusted EBITDA and CapEx as well as the fact that we do expect there to be an outflow of capital from working capital as we ramp up for R2.

Beyond that, as we look ahead to the Georgia site and facility, we do have a project-based finance loan with the Department of Energy that we anticipate being able to utilize for the continued build-out of our Georgia site and the tooling associated with the future vehicles that will be housed at that site. And beyond that, we'll remain opportunistic as it pertains to maintaining a strong balance sheet position for the future.

Okay. I want to give our audience one more chance to ask questions, we've got about 6 minutes left. If not, I've got more questions. Anyone in the audience have a last question for Claire? Don't be shy again.

Okay. Great. I'm going to keep at it. And Claire, thank you for enduring this. I appreciate it very much.

So there's been a relatively less conversation around Amazon. You just touched on it and Rivian's EDV offering. Can you update us on how the company is delivering on their electric fleet goals and remind us what the initial agreement was, if you can?

And then just beyond that, how do you -- I guess, I believe it's not exclusive is where I should be going with this. So maybe talk about the opportunity set there beyond just with Amazon?

Amazon's initial order was for 100,000 commercial vans by 2030. Part of the work that we've been doing alongside Amazon over the course of the last couple of years has been the build-out of infrastructure to support the electrification of the Amazon fleet. And they've put that foundation in place. And so as we look at our guidance for 2026, we do anticipate there being an increase in our volumes of delivery vans to Amazon, which is embedded within our guide for the year.

And we're also launching a couple of new variants for Amazon. So we're adding a larger battery pack unit and all-wheel drive configuration so that they can use the electric vans for more rural routes versus today where more of the commercial van deliveries in our products are being deployed in suburban and urban types of environments overall.

So Amazon is a great partner of Rivian's, and we're really excited about the continued deployment of this offering. There's roughly 30,000 vans as reported by Amazon on the road from our electric delivery vans for them. And then as we look to the future for other customers, we see there as an opportunity over the longer term to continue to educate large fleet operators on the advantages of an EV fleet.

And we see that through the total cost of ownership advantages that our EDV offers to Amazon and their appreciation of the efficiency and deployment of these vehicles within the fleet. But I will say that it does require a level of investment as Amazon has done in their charging infrastructure. And so that's an area where we're continuing to educate fleet customers. So don't anticipate significant volumes outside of Amazon in the near term, but remain confident that over the longer term, we will see more fleets electrify.

Terrific. I want to go back to Georgia, if I can. So recognizing that's going to create a ton of jobs, and that's obviously a positive, but we also have an administration that in various areas as it relates to grants and electrification has maybe pulled back a little bit. I'm curious, has anything changed in your conversations with the administration at all, and just how your outlook may have evolved, if at all, for the plan?

We're very aligned with the administration on the opportunity to bring more manufacturing jobs to the U.S. Rivian will be adding 2,000 jobs in our Normal, Illinois facility as we ramp up R2, and we'll be adding another 7,500 jobs in Georgia as well.

Beyond the job creation, Rivian is also a leader in the advancement of technologies as we think about the evolution of the automotive industry and the role that Rivian can play as a key enabler for the transition for many more traditional OEMs as well through the feature set and capabilities that Rivian can provide. So there's broad-based alignment as you think about our capabilities and the job growth and creation that Rivian can bring to market.

Okay. So the relationship is great. Everything is on track. Perfect.

Maybe my last question here as we're running out of time. I think at one point, Rivian was contemplating selling the R2 into Europe, which at the price point and given the market dynamics there broadly would seem to be attractive, at least from my perspective. Any updated thoughts on potential timing? And -- or is the focus just specifically at this point, getting the R2 right, getting the rollout right and then going from there?

So R2 is designed for global markets and not just to be a North American-centric product. We have large international ambitions as well as a company. And there's been a huge level of interest from a lot of Europeans that follow Rivian, that track the brand, part of the media reviews was through a couple of elite outlets out of Europe as well, all who had glowing reviews of the product. So we're really excited about the opportunity now that we'll have R2 in place to begin to build out the brand and integrate Rivian's across the continent of Europe over time.

Okay. That's a great stay tuned. A great place to stop. Claire, thank you so much. Rivian, thank you so much for being here. I appreciate all the time.

Good afternoon, and thank you for joining us for Rivian's Fourth Quarter and Full Year 2025 Earnings Call. Today, I'm joined by RJ Scaringe, our CEO and Founder; Claire McDonough, our Chief Financial Officer; and Javier Varela, our Chief Operations Officer. Before we begin, matters discussed on this call, including comments and responses to questions, reflect management's views as of today. We will also be making statements related to our business, operations and financial performance that may be considered forward-looking statements under federal securities law.

Such statements involve risks and uncertainties that could cause actual results to differ materially. These risks and uncertainties are described in our SEC filings and the shareholder letter we filed with the SEC. During this call, we will discuss both GAAP and non-GAAP financial measures. A reconciliation of historical non-GAAP to GAAP financial measures is provided in our shareholder letter. Just before the earnings call, we published and filed our shareholder letter, which includes an overview of our progress over the recent months. I encourage you to read it for additional details around some of the items we will cover on today's call.

Following our prepared remarks, we will be taking questions from sell-side analysts. In the interest of keeping the call to 1 hour, we would ask these analysts to limit any follow-on questions to one. With that, I'll turn the call over to RJ.

Thanks, Chip. Good afternoon, everyone, and thanks for joining us for today's call. 2025 was a year focused on execution at Rivian as we laid the foundations for scaling our business. Our team progressed the development of our technology road map in R2, while simultaneously driving continued improvement in our customer experience and our path to profitability.

In founding Rivian, I wanted to demonstrate how a clean sheet technology-focused vehicle could eliminate long accepted compromises and provide consumers choice. Our goal with the launch of our R1 products was to establish the Rivian brand by delivering a combination of efficiency, on-road performance, off-road capability, functional utility and product refinement that simply didn't exist in the market. The first vehicles established Rivian as a brand that enables people to do the things they love, enable adventure as well as transcend different segments, form factors, use cases and geographies.

In the fourth quarter of 2025, the R1S was the best-selling premium electric vehicle priced above $70,000 in California, New York, New Jersey, Oregon, Virginia and Washington, D.C. And it was the best-selling SUV EV or non-EV, over $70,000 in the state of California. Now I'm excited that we are months away from starting customer deliveries of R2, our first mass market vehicle. One of the things that's often overlooked around EVs is that there is a surprising lack of high-quality choices at prices around or below $50,000 for a new vehicle in the United States, there are only a few compelling EV choices as compared to hundreds of internal combustion or hybrid options that have a wide range of form factors and design aesthetics.

From the lens of the customer, if you want to buy a midsized SUV with robust technology, autonomous capabilities and a reasonable price point, you've really only got one choice, and it's been that way for a long time. This is a reflection of a market that's being underserved. We believe R2 is going to change that. R2 is an extension of the experience we delivered in R1 with design elements and performance to inspire adventure, but in a smaller form factor and importantly, at an attractive lower price point. Launch Edition R2 variants will be well equipped with a dual motor all-wheel drive setup that provides more than 650 horsepower and over 300 miles of range. In mid-January, I was thrilled to drive our first R2 manufacturing validation build off the production line in our factory in Illinois.

As you've seen from the extremely positive, media reviews of our preproduction vehicles over the last few days, R2 is an exceptional vehicle, and I believe will be a game changer for our customers, our company and the industry. One reviewer said, the R2 is an exceptional vehicle, quite possibly the best all-around electric vehicle I've ever driven. We look forward to getting investors and more media in R2 for demo drives so they can experience the capabilities of the vehicle. We plan to provide additional product, pricing and lineup details on March 12. Turning to our AI and Autonomy Day. It was great to see so many of our stakeholders at our offices in Palo Alto this past December. We were excited to showcase our innovation across our vertically integrated hardware, software and autonomy teams and unveil RAP1. Developing our own chip was driven by the need for velocity, performance and cost efficiency and is a key development of our autonomy platform.

Near the end of last year, we released universal hands-free, which expanded our advanced assisted driving capabilities for Gen 2 customers to more than 3.5 million miles of roads across North America. Since its release, customer utilization of our autonomy features has doubled. Rivian is also making significant progress in making software and AI core to everything we do from the way we design, develop, manufacture and service our cars to the way our customers interact with their vehicle. This is enabled by the Rivian Unified Intelligence, a common AI foundation that understands our products and operations as one continuous system and personalizes the experience for our customers. It also defines how applications will integrate in our vehicles in the future. We were excited to demo the Rivian Assistant at AI and Autonomy Day and expect to launch this feature early this year.

Finally, we continue to see the extensibility of our electrical hardware and software platform with the work happening in our joint venture with Volkswagen Group. I'm very pleased that we have delivered vehicles for winter testing for multiple Volkswagen Group brands, 13 months after the formation of the joint venture. In closing, 2025 was a foundational year for scaling Rivian, and I could not be more excited for the year ahead. I believe 2026 will be an inflection point for our business. As an American automotive technology company that develops and manufactures incredible electric vehicles, we believe that the future of the automotive industry will be fully electric, autonomous and AI defined. I've never been more confident in the opportunity ahead for Rivian than I am today.

I firmly believe Rivian's technology, along with our direct-to-customer ownership experience, position our company to build a category-defining brand with a strong mass market product portfolio for the U.S. and global markets. With that, I'll pass the call over to Claire to discuss our financial results.

Thanks, RJ, and good afternoon, everyone. As RJ discussed, we believe 2026 will be an important year as we scale our business. Launching R2 will extend our brand to the mass market, and we expect R2 will drive meaningful automotive segment growth and profitability over time. Now before I dive into the quarter, there are a few key financial metrics that I'd like to highlight for 2025. First, on a full year-over-year basis, we delivered nearly $5,500 of improvement in average sales price due to the introduction of our second-generation R1 quad models, a higher mix of R1 units and increased base prices for the 2026 model year.

Second, on a full year-over-year basis, we achieved an approximately $9,500 improvement in automotive cost of goods sold per unit due to material cost reductions and operational efficiencies. Finally, the improvement in unit economics in our Automotive segment, when combined with our strong software and services performance, resulted in greater than $1.3 billion of improvement in full year gross profit, making 2025 our first full year of positive gross profit.

Additionally, our gross profit performance, coupled with our focus on cost management, enabled our adjusted EBITDA for 2025 to be at the favorable end of our guidance. All of these metrics represent our continued progress in the operational efficiency and profitability of our business, which sets a strong foundation for 2026 and beyond. We expect the gross profit per unit for R1 and the commercial vans to be further enhanced as we ramp up production and deliveries of R2, coupled with the gross profit contribution of R2 over time. Turning to the results of the fourth quarter. Our consolidated revenues were approximately $1.3 billion.

Consolidated gross profit was $120 million, and our gross profit margin was 9%. Gross profit included $108 million of depreciation and $26 million of stock-based compensation expense. Adjusted EBITDA losses for the fourth quarter were negative $465 million, $137 million improvement from Q3 2025 due to higher gross profit and lower operating expenses. Now looking at our Automotive segment. During the fourth quarter, we produced 10,974 vehicles and delivered 9,745 vehicles from our manufacturing facility in Normal, Illinois. This was the primary driver of the $839 million of automotive revenue. Automotive gross profit for the fourth quarter was negative $59 million, a $71 million improvement from Q3 2025 due to a higher mix of commercial vans, which resulted in the lowest cost of goods sold per unit in our history.

Our Software and Services segment reported another strong quarter with $447 million of revenue and $179 million of gross profit. $273 million or approximately 60% of software and services revenue was attributable to our joint venture with Volkswagen Group. We also experienced strong growth from our marketing and vehicle repair and maintenance. Looking at our balance sheet, we ended the year with approximately $6.1 billion of cash, cash equivalents and short-term investments. In 2026, we expect to receive an additional $2 billion of capital as part of our joint venture with the Volkswagen Group. $1 billion is an investment subject to the successful completion of winter testing, which RJ discussed earlier and $1 billion is nonrecourse debt, which we expect to receive in October. Finally, for our 2026 guidance, we expect to deliver between 62,000 and 67,000 total vehicles across R1, R2 and our commercial vans.

We expect total deliveries of approximately 9,000 to 11,000 per quarter in the first half of 2026. We plan to start production of the R2 launch variant with a single shift and expect to add a second shift towards the end of the year. While we believe our gross profit will increase year-over-year, we expect the complexity of a new vehicle launch to negatively impact our automotive gross profit in the second and third quarters before becoming a benefit to our overall operations in the fourth quarter as we ramp production and deliveries. As a reminder, we believe this is a transition year for the Automotive segment's profitability.

Delivering a strong exit rate for R2 production and deliveries will be a key focus for our team. For 2026, we expect an adjusted EBITDA loss of between $2.1 billion and $1.8 billion. Our adjusted EBITDA guidance also includes a step-up in R&D spend as we accelerate investments in our autonomy road map and look to deliver LiDAR, our first RAP1 chips and limited point-to-point functionality for our customers by the end of the year. We believe autonomy will be a key fundamental long-term differentiator for our business.

We also plan on the continued growth of our SG&A, driven by the expansion of our service and sales footprint as we scale with R2. Finally, for 2026, we expect capital expenditures of $1.95 billion to $2.05 billion related to finalizing construction and tooling for R2 and normal, kicking off vertical construction for our greenfield plant in Georgia and the continued build-out of our sales, service and charging infrastructure. In closing, thank you again to the team for delivering a great 2025.

As we look forward to 2026, we remain steadfast in our belief that R2 and our technology road map will be truly transformative for our growth and profitability. I'd like to turn the call back over to the operator to open the line for Q&A.

[Operator Instructions]. Our first question comes from Emmanuel Rosner with Wolfe Research.

My first question is on the cadence that Claire, you just highlighted with the 9,000 to 11,000 units per quarter in the first half and then obviously, 62,000 to 67,000 on a full year basis. Is that 10,000 per quarter or so, is that your expectation for R1 plus EDV for this year in 2026? And then the upside in the second half would be essentially delivering the R2?

Thanks, Emmanuel. As you think about the cadence, as RJ articulated in his prepared remarks, we expect first deliveries to begin for R2 in the second quarter. Like any ramp, the number of deliveries will be rather small as you think about the Q2 impact of R2 contribution to the 9,000 to 11,000 units per quarter that we anticipate in the first half of the year.

And then as we get into the second half of the year, we expect to see the continuation of the ramp of R2, coupled with the ongoing deliveries of our commercial van as well as R1. So on a full year basis, you can think about the R1 commercial van being roughly in line with our 2025 total volumes.

Okay. And then another one on the cadence side, but more from a financial point of view, I think in the past, you had targeted, I think, by the fourth quarter of this year, some level of profitability on R2 to sort of like demonstrate essentially the potential. Is that still the expectation for this year?

Yes. As I mentioned, we expect 2026 will be a transformational year for our automotive gross profit, and we expect that both R2 [Audio Gap].

Our next question comes from Dan Levy with Barclays.

First, I wanted to start with just a question on the R2 volume assumptions, which you just talked about a moment ago. As you're going into the launch, do you have a feel for what the aggregate demand is? And maybe you could just talk to the question of your confidence on people wanting to take delivery of R2 before the new ADAS platform or hardware is put into the vehicle? How much of a current the lack of that new hardware will be on deliveries?

Dan, thanks for the question. With regards to R2, I've had chance to spend a lot of time in it over the last several months and really over the last month or so, driving our validation vehicles that are produced in our plant off of our manufacturing validation build. And the vehicle is just absolutely incredible. It's the combination of features, the packaging, the vehicle dynamics, the steering wheel, we're incredibly bullish on. And as I talked about a lot in the past, we ultimately think the market is really hungry for some choice in this segment.

As I said in my opening remarks, just the lack of choice that exists in and around this $50,000 price point has led to very high market share concentration of one vehicle. And so this is the first time there's going to be a real alternative. And this is important for folks that are in internal combustion vehicles today, midsized SUVs and looking for something that fits their form factor needs, their aesthetic needs, their packaging needs. And so we're very, very focused on putting this together. And with that said, we have a lot of confidence in the overall demand of course, that's why we leaned so much into the program and lean so much into what we're about to launch.

And just the issue of the ADAS platform, people taking it before the new hardware comes in?

Yes. With new technology, there's always -- especially for us as a business, given that we've got an enormous focus on developing new technology, there's always something new coming and recognizing that there are a lot of customers that are just waiting to get a great midsize SUV. And so given the enormous backlog of demand we have as well, the short period that we have where we'll be launching with essentially an upgraded version of our Gen 2 autonomy stack before our Gen 3 autonomy stack comes -- we don't think that's going to be a significant issue for those that want to wait for it, they certainly can and for others that are going to be excited the vehicles in a short time frame, be available prior to that.

Okay. Great. As a follow-up, I wanted to ask about partnership or licensing deals with other automakers. And maybe you could just give us a sense of the tone or tenor of discussions on licensing the network architecture to others. And you said at your Autonomy Day that one of the opportunities on your in-house processor was not only that it could be used for your vehicles, but you could also sell this to others. It's a better, what you said, bang for buck. So I know it's early days, that still isn't out and the initial units of that process are going to need to be for you. But what types of discussions are you having with automakers on potentially selling that or licensing that to them?

So as I said in the opening remarks, we've had in the first 13 months of establishing a joint venture with Volkswagen, we've testing on multiple VW Group products. And I think the true demonstration and existence proof, if you will, of the scalability of our technology in terms of being able to work across multiple form factors, different price points, different brands and importantly, be productized into a platform that can go across a large existing OEM. This relationship is really important for that. And so that's our focus. But of course, we have relationships with a broad spectrum of other manufacturers. And I've said this in the past, but I deeply believe that over the course of the next several years, every manufacturer has to make the decision as to whether or not to get to a software vehicle or they're going to develop it themselves, secure it from a third source of which we will be the only demonstrated example of having scaled this technology outside of our own products or accept that without the technology, you will lose market share. And so we're quite bullish on the potential for this technology platform, and we see it as really an important part of our portfolio going forward.

Ladies and gentleman, we are having further technical issues, please standby, we will resume the Q&A as soon as possible. Thank you. [Audio Gap].

Our next question comes from Ben Kallo with Baird.

Congrats on all the progress so far. Maybe first, can we talk about just the VW relationship and there was an uptick in revenue from that. And just if you give us a sense of how that progresses and the potential to expand the relationship or how we should think about it growing? And then the second question is just around Georgia, the DOE loan guarantee, how you're looking at that and maybe just liquidity in general as you start working on that.

Thanks, Ben. And I spoke a little bit about this in the previous response, but I understand we've had a few technical difficulties here. So at risk of repeating myself, I'll just say -- cover a few things that I said earlier. The Volkswagen relationship continues to progress. We're now 13 months since the joint venture started. We've -- with that, we've started winter testing on several different Volkswagen Group products. And of course, we're working towards the first launch of those vehicles in 2027.

And the relationship has been very, very strong. We had a great session, in fact, coincidentally last week with a broad set of the Volkswagen Group leadership team. And seeing this be used and deployed across not just vehicles of similar price point to Rivian's vehicles, but across the price point -- across a wide band of price points and across a range of form factors is really important. And it really demonstrates the scalability of the technology. And so ultimately, we're going to continue working towards delivering multiple Volkswagen Group products, but this does, of course, open the door for opportunity with other manufacturers as well.

And then to put some of the financials behind the software and services outlook as a whole, we anticipate seeing that we'll approach about 60% year-over-year growth in our software and services business, and it will be a significant driver of our gross profit outlook as well with margins that we expect to be in the mid-30% area as a whole. Then as your second question, which was on the capital road map, as I mentioned in my prepared remarks, we ended 2025 with $6.1 billion of cash, cash equivalents and short-term investments. We expect that we'll receive another $2 billion from Volkswagen Group throughout the course of 2026. There's still roughly another $0.5 billion payment associated with the original joint venture transaction as well that will happen, we anticipate in 2027.

And then as it pertains to our broader capital road map, we'll continue to remain opportunistic as well on that front. On the DOE loan question, similar to what we've shared in the past, we certainly share the President's desire to bring jobs back to the United States. We're excited to keep up our work on creating new American manufacturing jobs. We'll be adding approximately 2,000 new jobs at our Normal, Illinois plant for the ramp-up of R2, an additional 7,500 jobs at our future Georgia plant as well. And similar to the comments RJ just made, Rivian is working to help drive innovation and technology leadership in the U.S. automotive industry for consumers and also associated with our joint venture with Volkswagen Group, enabling this technology for the industry as a whole.

Our next question comes from George Gianarikas with CG.

As it relates to your guidance for vehicle sales this year, to the extent you see a strong conversion in your backlog for the R2, could you see upside to that? Or are there certain production bottlenecks that you'll have to work through towards the end of the year?

Thanks, George. The process of ramping a vehicle is something we've spent a lot of time talking about in this forum, but certainly internally, we're really putting a lot of effort on making sure we have a very smooth production launch and then associated ramp. And we often think of the plant as being the bottleneck for the ramp. But in fact, we have to remember there's hundreds of other companies that are providing components into Rivian that ultimately really contribute and are a key part of the ramp.

And ramping our supply base is something that we're very focused on and planning around. And you can only ramp as fast as your slowest part, so to speak. So with that said, we've -- as you heard earlier, we're starting with a single shift. We're bringing on a second shift that will be happening near the end of the year. And then in 2027, we'll be adding our third shift. And this has been very methodically laid out to make sure that we're ramping consistently and evenly across the supply chain. And so certainly, and as you alluded to, there's going to be a large demand backlog that we're working through and a tremendous amount of excitement for the R2 vehicle. And as more reviews come out about it and more people get exposure to it, we expect that to continue to expand and grow. It's worth noting some of the preproduction reviews that came out last week, the feedback has been universally super positive. And so we're acutely aware of that, but we are working very carefully to coordinate the ramp and coordinate the growth of output with our supply base.

And maybe as a follow-up, as you sort of crystallized your selling prices and your cost structure for the R2, can you just maybe speak to the guidance you gave at your Analyst Day last year about reaching EBITDA positivity in 2027 and your long-term vision around having 25%-ish gross margins and high teens EBITDA margins?

Thanks, George. As prepared -- as discussed in our prepared remarks, we believe that 2026 is going to be a transition year for the automotive gross profit segment. And our North Star for the normal plant is going to be getting our production and deliveries up to about 4,000 units per week. While there's a lot of execution required, as RJ just walked through from the team in order to achieve that outcome, if we're successful, we believe it would put the company in a strong position to achieve our adjusted EBITDA goals. And as we look at the broader outlook, we certainly see there being 20% gross profit target for our automotive segment. As we think about the overall contribution of software and services, there's many outcomes or licensing deals that [ Rubin ] could do that could allow gross profit to be much higher than the 25% in our end state as well.

Our next question comes from Joseph Spak with UBS.

Just a couple of questions. Claire, I know you said the second shift starts in the back half. Does the guidance contemplate like exiting the year at a full 2-shift rate for the R2? I just want to sort of understand how we should think about a jumping off point for '27.

Sure. As you can imagine, Joe, we'll be in the process of ramping up the second shift. So as you think about the exit rate of 2026, we won't yet be at full production across 2 shifts. We'll be getting there as we continue to progress all of our operational efficiencies and get the team ready to ramp up throughout the course of 2027.

And then I guess, RJ or Claire, just you mentioned more details on March 3. I'm assuming that's also when reservation holders will be invited to configure. But is there any update you could give us on that order book? And then on the pricing side, one of the things we've seen, again, not Rivian specific just to the industry and the world really is the cost side between metals and memory. So curious to sort of wonder how you're thinking about that impacting either your pricing for the vehicle or whether that's contemplated in the EBITDA guidance for the year?

Yes. So on March 12, we'll be providing the full picture around the overall portfolio of products that will exist for R2, and that will include the launch configuration, which, as I've said, is a dual motor performance variant with a premium trim. The different combinations of trim, powertrain performance and battery size are what we'll be describing in detail on the 12th. And then along with that, allowing customers to start configuring their vehicles and allows us to start getting ready to be making deliveries later in Q2.

I think important here, as you think about the overall demand profile for the vehicle, we've put a lot of time into thinking about the different combinations and recognizing what different folks are going to want and having the benefit of having lots of conversations and also the benefit of seeing what are the most popular trims and configurations in R1. And so we're really excited to go through that, but it's something that we do want to go through as -- present it as the full meal as opposed to giving little bits and pieces. So other than talking about the launch configuration, the rest of the configuration is going to be something we talk about in detail in about a month.

And then Joe, our adjusted EBITDA guidance does contemplate some of the increases that we've seen in raw material costs and the current supply chain backdrop as well.

Our next question comes from Mark Delaney with Goldman Sachs.

I was hoping to speak more about automotive COGS. Maybe you can help us better understand what led to the reduction in cost per vehicle, both sequentially as well as year-over-year. And then as you look forward, clearly, you just alluded to some of the supply chain challenges around DRAM and other input costs that are embedded into your guidance. But where do you ultimately think the R2 cost can get to? And is the 50% reduction in the BOM cost compared to R1 still the right level to think about?

Sure. For Q4, we were able to deliver $92,000 of COGS per unit, and that was about a $4,000 per unit improvement relative to the third quarter. One of the key drivers was associated with the mix shift. So we had higher penetration of commercial vans in the fourth quarter relative to the third quarter. And then beyond that, the ongoing operational efficiencies that we continue to execute across our normal operations as well that also contributed to the reduction in our COGS per unit. As we look at the full year-over-year improvement, the biggest driver that we saw was associated with the reduction in our material cost, that was both the transition for us to move fully to Gen 2 vehicles.

We also, in addition, saw a significant step down in terms of a lot of our raw material costs and importantly, a step down in the cost of our lithium prices that was another contributor for us as well as we looked over at the full year-over-year step down in terms of COGS per unit, coupled with the ongoing operational efficiencies that we continue to progress throughout the course of the year. As we look forward to R2, one of the key factors for us with R2 is the opportunity for us to have also over the course of the last year, started to see some meaningful benefit from our joint venture -- joint sourcing opportunities associated with our low-voltage electronics that we're sourcing for the R2 vehicle.

And that's really been a key enabler for us to continue to progress the material cost trajectory of the R2 product as well. And then as we approach the Georgia plant, we'll have further opportunity to drive additional synergies or efficiencies as we start to source future vehicle volumes as well that will share the fundamentals of the midsized platform as a whole. So those are a couple of the puts and takes. The other callout that I would highlight as well is we do anticipate seeing a reduction in terms of our tariffs per unit. So we didn't have the full benefit of the Section 232 offsets for the entirety of the fourth quarter. So we'll see further benefit from a tariff per unit standpoint as we progress forward, and R2 will also benefit from that in the future.

My other question was on EDV, and I understand probably flattish volumes there for 2026 based on your comments earlier on the call, but you did speak to a plan for some additional variants, including one with more range for Amazon. So maybe help us better understand when to expect that new product for the commercial segment and what that might mean for van deliveries and the broader commercial opportunity going forward?

We do expect some growth in our EDV demand in 2026. And as you called out, there's an all-wheel drive version of the van and a larger battery pack variant as well. And both of those are to help unlock specific use cases within the Amazon network. We're working really closely with Amazon in defining the requirements of those and excited to get those launched. And the relationship with Amazon continues to be very positive. And certainly, the EDV continues to perform extremely well.

Our next question comes from Chris Pierce with Needham.

As you talk about adding the second shift, then adding the third shift, can you kind of just walk through setting up hiring for these workers in the normal area? I'm just not sure if we should think about that as a barrier or a burden or just kind of -- is it already in motion? Or do you already have these and you kind of move people from one shift to another? Just kind of any detail around that, please?

Yes. Chris, thank you for your question. Hiring process is in place, is proceeding according to plan. We have enough candidates. At the beginning, it was even spontaneous candidates that wanted to work for us. So we are good from that end. There's part of the team that will populate the R2 line that is coming from the existing flows, but it's an important part as well that is hired from outside.

We have reinforced our training programs. We have even before hiring the people pre-hiring activity, just to let them know what is working in the lines and what they should expect there. And so far, we are very happy with the response of the talent pool and the people pool that we have seen there. So...

Perfect. And then just -- I guess, I just kind of want to understand, you've had the reviews this week. You're kind of flipping over the card as far as first trim and other trims in a month. And then we could see initial deliveries sometime in the second quarter. I guess I'm just kind of curious I'm just thinking about maybe Apple and the iPhone or the iPad, and it's just sort of boom, here it is, you can buy it now. I guess I'd just love to hear about why the spacing? Is there a psychological effect or it just comes down to what you can produce, when you can produce it at the plant, but just the timing of the cadence as you move towards the launch.

Well, thanks, Chris. One of the amazing things about the R2 program is there's an enormous backlog, but that does create a challenge for us with making sure essentially how do we select who receives our vehicles first and having a processor on that. And so opening up the reservation or opening up the configuration process allows us to start taking this demand backlog and organizing around when we make deliveries and who gets the vehicles first. It's not as if you could like press a button and instantly have thousands and thousands of vehicles available.

So we start producing, we are ramping production, but the demand will outpace our ability to produce. And so that process allows us to organize in a thoughtful way and learning a lot from some of the past launches we've had, how do we prioritize and how do we sequence deliveries to our broad base of customers.

Our next question comes from Itay Michaeli with TD Cowen.

I wanted to actually ask on the universal hands-free. Curious how initial feedback has been since December, how we should think about feature improvements and OTA updates this year? And maybe what you're also assuming for paid subscriptions this year?

We're really excited to talk about this. This is a huge effort within the business and autonomy and AI was, of course, focused on all the work that we're doing here. But our Universal Lands Free is really think of it as the first step in a whole series of steps that expand the capability. And so Universal Lands Free expanded the number of miles where you can drive with hands off the wheel, but eyes on the road. It's around 3.5 million miles, essentially any road with marked lanes. And later this year, we'll be unlocking the ability or enabling the ability for the vehicle to drive point to point. So you put your address into the vehicle and the vehicle navigates to that address.

And then the next steps ultimately are driving towards what we think of as personal Level 4. But between point-to-point and personal Level 4, we'll have hands-off eyes off, so you'll be able to take your eyes off the road and do other things starting first on the highway and then expanding from the highway. And then following that, we'll have our first Level 4 applications within a geofence area to start, but ultimately expanding over time.

And our view and really strong conviction is that over the course of the remainder of this decade, we're going to see autonomy go from something that today with hands-off capabilities certainly is a nice feature to have. But as we start to move to hands-off and eyes off and then ultimately to Level 4 where the vehicle can operate itself entirely on its own, including driving empty, it really creates a whole new customer experience, and we think it becomes a critical part of the purchase decision. And this will -- this is going to drive significant change in how we think about the business model. It's going to drive significant change in how consumers think about what vehicles they want to purchase.

That's very helpful. As a quick follow-up on the financials, any sense of how we should think about working capital flows this year, particularly as you go through the initial R2 ramp?

Sure. We will see working capital be an outflow of cash for us over the course of 2026. And that in part is driven by the buildup of our inventory balance associated with the launch of R2.

Our next question comes from Andrew Percoco with Morgan Stanley.

I actually just want to come back, RJ, to what you just said on autonomy really driving the value proposition and driving experience over the next few years. And I guess I'm just curious, like can you share your thoughts around whether or not there will be a retrofit opportunity for existing or new R2 customers that don't have Gen 3 with LiDAR and existing R1 customers? And any thoughts of when you'll introduce LiDAR onto R1 production?

We're not -- in terms of retrofit, this isn't something that's contemplated or planned. Certainly, there will continue to be over their updates for our Gen 2 vehicles, our R1, Gen 2 vehicles and our launch R2 vehicles. But in terms of hardware upgrades, those are not planned. The hardware upgrades that we've talked about in the past, we talked about at our autonomy and AI Day, these are going to be on vehicles in the early part of 2027.

And certainly very, very excited about those. But the capabilities and even the demo that we had at our Autonomy AI Day, which was a point-to-point demo, that's being done. That demo was on a Gen 2 of our R1 vehicles. So that's -- that will be available on any Gen 2 R1 vehicle as well as the R2 vehicles we'll be launching with.

Got it. Okay. And can you get to -- is the plan to get to eyes off point-to-point with Gen 2? Or is that something that's going to require Gen 3?

I think the important thing to keep in mind on our upgraded architecture, which is, as I said, coming in '27, is that has a few really important purposes. One, of course, is it raises the ceiling on what's possible. So it grows the opportunity to add even more capability beyond point-to-point, but it also serves as an even more enhanced part of our data flywheel where we have enhanced cameras, higher level of inference in vehicle, but importantly, we add a LiDAR, as you referenced, which turns essentially every vehicle into part of our ground truth fleet, which is really helpful for training our end-to-end model.

And so the way that the model continues to improve is we're benefiting from the thousands and thousands of drivers that are on the road and the vehicles that are on the road, pulling interesting and unique events back off the vehicles and allowing us to feed that into the overall training loop that we have for what we call our large driving model.

Got it. Okay. And just one quick clarification. When you say Gen 3 will be available early 2027, will that include R1? Or is that just still R2?

This is for R2.

Our next question comes from Edison Yu with Deutsche Bank.

This is Winnie on for Edison. Can you guys hear me?

Yes.

My first question is on the relationship with VW as it matures. I was wondering if the topic of how to best utilize vehicle data come up. And more asking this in the context of VW naturally having a much larger fleet and Rivian being able to potentially benefit from getting access to that data for training. So just curious your thoughts on that topic.

The Volkswagen fleet and what ultimately we're delivering to Volkswagen from a technology platform doesn't include our autonomy platform. So it's our embedded software platform, our topology of ECUs, including our zonal architecture, but it doesn't include our self-driving architecture.

Got it. And then maybe just on the RAP1 chip. Curious to hear your maybe longer-term inspirations for that. Is this something you think that will be limited to Rivian? Or do you envision maybe this being used by some of your partners or other OEMs and whether it could be potentially applied to nonautomotive products like [ humanoid? ]

I guess I'll answer it really broadly. This is -- when we think about our self-driving and autonomy efforts, this is an enormous focus for the business and represents our most significant area of capital investment from an R&D point of view. And so when we look at this through the lens of the next several years, we do believe that as we continue to demonstrate progress and work towards that growing capability set that I talked about before, ultimately culminating in Level 4, this is a platform that certainly beyond Rivian has applications. And so we do envision a world in which this becomes something that we can monetize through a few different ways.

We can monetize it through increased market share and growing number of vehicle sales. We can, of course, do that through new and unique business models and new ways of thinking about consuming transportation, and we can do that through selling and providing the technology to other manufacturers. Now specifically to our RAP1 processor as well as its future variants of that processor, we do see applications for vision-based robotics well beyond the vehicle. Of course, the vehicle is a great near-term vision-based robot. But even within Mind Robotics, which is a new company that we created, that is also an example of a great customer application and a great use case application for our RAP1 processor.

Our next question comes from Tobias Beith with Rothschild & Co. Redburn.

May I ask what Rivian's management's latest thoughts are on captive battery cell manufacturing and assembly are considering the recent development in the price of lithium salts. I know that this activity was contemplated at one point at your forthcoming plant in Stanton Springs.

Well, thanks, Toby. We've found great working partnerships with our battery cell suppliers and have taken a very active role in securing and sourcing some of the upstream precursor materials. So you called out lithium. That's a really great example of an area that we, from a sourcing point of view, spend a lot of time on. But being able to work with these key battery partners and leverage the investments they've made in production capacity and in their own cell construction and cell design has been really helpful for us in terms of efficiently deploying capital.

Our last question comes from James Picariello with BNP Paribas.

Can you hear me?

Yes, James. We can hear you.

Great. [ Vista ] curious on your thoughts regarding R1's potential to see additional demand from one of your major competitors announcing the end of its 2 high-end models next quarter, right? We're talking about 20,000 annualized units essentially up for grabs in the U.S. It just seems like the R1 can be a natural home for many of those buyers. Just curious your thoughts.

Yes. The Tesla Model X, which, as you said, will stop production, at least what's been said is next quarter. The Model X is a really important product from an electrification point of view, and it was one of the first more at-scale products to show customers how exciting electrification can be. And so as that leaves the market along with the Model X, it does create an opportunity. There's -- I've talked about this a lot in the context of R2, the lack of choice, but this also is true in the case of R1. And R1 as it stands, our R1S is -- I called out in my opening remarks, it's an enormously successful product in this premium price category.

So it's the best-selling premium SUV, electric or nonelectric in the state of California. And then it's the best-selling premium electric SUV in the United States and the best-selling premium electric vehicle, SUV or non-SUV in a number of states across the U.S. And so with even less choice now in that price category, it does represent an opportunity for us. But I'd just say very broadly that the overall lack of choice, and this is really true in the price category of R2, we think is an enormous opportunity for us to capture market share and, of course, provide something really exciting to customers.

Yes, makes a lot of sense. My follow-up, just for the strong gross profit contribution slated for this year, can you help dimension at all what the expected contribution might look like from VW as we think about that 60% year-over-year growth?

Sure. As you think about the Software and Services segment, roughly half of our revenue comes from revenue streams associated with our joint venture with Volkswagen Group, and we expect that to largely remain true as we look ahead to 2026 as well. And it is a more disproportionate share of the overall gross profit dollars that we earn out of the [ Safran ] Services segment as a whole.

This concludes the Q&A section of the call. I would now like to turn the call back to RJ Scaringe for closing remarks.

Well, thanks, everybody, for joining the call, and we apologize for some of the technical difficulties that we had at the start. But -- we are -- hopefully, a takeaway from this is just how excited we are about R2. As I said, I've had a chance to spend a lot of time in the car and in a variety of ways, whether it's taking my kids to sports games or loading up the back with gear, and it is just such an incredible embodiment of the Rivian brand and captures so many -- so much of the essence of what makes R1 a special vehicle, but at a price point that, as we said, starts at $45 and will allow a much larger number of customers to access the vehicle.

And so as we think about the next couple of months, we, as a company, remain incredibly focused on the launch ramp of this vehicle, along with the continued development of the technology we're building, both on the software side, which we talked a bit about, but certainly on -- also on our autonomy side. And so with that, thanks again for joining the call, and we're looking forward to a lot of folks being able to experience the R2 themselves.

Please welcome RJ Scaringe, Founder and Chief Executive Officer of Rivian.

We are incredibly excited to host everybody here today. We are in Palo Alto, which is the hub for Autonomy and our technology development teams. AI is enabling us to create technology and customer experiences at a rate that is completely different from what we've seen in the past.

If we look forward 3 or 4 years into the future, the rate of change is an order of magnitude greater than what we've experienced in the last 3 or 4 years. Directly controlling our network architecture and our software platforms in our vehicles has, of course, created an opportunity for us to deliver amazingly rich software. But perhaps even more importantly, this is the foundation of enabling AI across our vehicles and our business.

I'd like to talk about Autonomy first. The field of autonomy really started about 20 years ago. And about, let's say, up until the early 2020s, the approach was centered on a rules-based environment where a set of perception sensors would identify and classify objects and hand those classified and vector-associated objects to a planner that was built around a human-defined rules-based framework.

A few years ago, it became clear the approach to autonomy needed to shift. With innovations around transformer-based encoding and the design of large parameter models, the approach has moved to building a neural net-like understanding of how to drive instead of following a classical rules-based approach. Recognizing this massive shift in how we approach autonomy, in early 2022, we began the process of a clean sheet design to our platform. This first embodiment of this work was in our Gen 2 R1 vehicles, which we launched in mid-2024.

With this updated platform, our Gen 2 vehicles now have 55 megapixels of cameras, 5 radars and run on an inference platform that was a 10x improvement over our Gen 1 vehicles. This platform was designed around an AI-centric approach. And with the deployment of our Gen 2 R1s, we began the process of building our data flywheel to grow and build our large driving model.

Now because this AI-centric approach represents a model trained end-to-end to the millions and millions of miles driven on our vehicles, enhancing the perception platform or improving the compute is accretive to the capabilities of the model, meaning the model only continues to get better as the perception and as the compute platform improve. So you can sort of think of it like this.

If you learn to drive and you had bad vision and suddenly you're able to put on glasses and see much better and then go even further and supplement that with new perception modalities of radar and LiDAR and then your compute here metaphorically [ brain ] was expanding in capability by an order of magnitude, you wouldn't forget the things you learned, the rules of the road, how to operate a vehicle, but your ability to understand nuance, to respond to complex situations and perceive the world in environments with poor or limited visibility would improve dramatically.

Our approach to building self-driving is really designed around this data flywheel. We're a deployed fleet, has a carefully designed data policy that allows us to identify important and interesting events that we can use to train our large model offline, before distilling the model back down into the vehicle.

Now while our R1 Gen 2 vehicles and the demos some of you are going to be trying today are using our Gen 2 sensor set and of course, the associated compute platform, over the last few years, we've also been developing our Gen 3 substantially enhanced platform, and this will underpin a massive leap forward with R2. Starting in late 2026, our Gen 3 Autonomy platform will include 65 megapixels of cameras, a robust radar array and a front-facing long-range LiDAR. And at the core of this platform, our first in-house Rivian Autonomy Processor.

Now the first iteration of our in-house inference platform includes a neural engine with 800 TOPS. It's optimized to support camera-centric AI in the physical world and enables a dramatic expansion of Rivian's Autonomy capabilities. When integrated into what we call our Gen 3 Autonomy Compute platform, which Vidya will go through later in detail, it will deliver 1,600 TOPs. The effectiveness and efficiency of our in-house processor has been a core focus in its development. Our Gen 3 computer is capable of processing 5 billion pixels per second.

Now you've heard me say this a couple of times. We've designed this entire architecture around an AI-centric approach where the data flywheel of our deployed fleet helps the model get better and better through reinforcement learning. Not only does this sensor set enable a much higher ceiling than what we have in our vehicles today, it also makes the platform much better to serve in building our model. We're going to continue to see improvements on our platform.

Later this month, we'll be issuing an over-the-air update to our R1 Gen 2 customers, and that will dramatically expand the existing hands-free capability, going from less than 150,000 miles of roads to more than 3.5 million miles of roads in North America. And again, this is just a step in a series of steps. Starting in 2026, we'll begin rolling out point-to-point capabilities in which the vehicle can drive address to address. So what that means is you can get into the vehicle at your house, plug in the address to where you're going and the vehicle will completely drive you there.

Now some of you are going to experience this today in our R1 vehicles. And of course, this will be embedded in R2. Now as I described, the Gen 3 hardware architecture launching in 2026 expands the ceiling of what we can achieve. The next major step beyond point-to-point will be eyes off, meaning you can navigate point-to-point with your hands off the wheel, but importantly, your eyes off the road. And this gives you your time back. You can be on your phone or reading a book, no longer needing to be actively involved in the operation of the vehicle.

And following eyes off, the next major step will be personal Level 4. And with this, the vehicle will operate entirely on its own. This means it can drop the kids off at school, it can pick you up from the airport. It allows it to really be integrated into your life in ways that we can't even imagine and haven't seen yet today.

Now while our initial focus will be on personally owned vehicles, which today represent a vast majority of the miles through the United States, this also enables us to pursue opportunities in the rideshare space. Now beyond self-driving, we've also created what we call the Rivian Unified Intelligence. This AI backbone exists across our vehicle and across our entire business. And we've talked for a long time about software-defined vehicles, which are really the foundational building block for an AI-defined vehicle, where every part of the vehicle, the experience and everything that's happening across the vehicle is designed around AI from our Rivian Assistant to enabling our direct-to-consumer sales and service model as well as our future manufacturing infrastructure.

Now as I said at the start, we could not be more excited about what we're building, and we have a lot of details to show you here today. And with that, I'm excited to introduce Vidya to talk about our hardware platform.

Thank you. My name is Vidya Rajagopalan, and I lead the electrical hardware team here at Rivian. My team is responsible for electrical content in the vehicle, ranging from our in-house 5-nanometer silicon that you just heard about that operates at voltages below a volt to the power electronics for electric motors that operate at 400 volts and lots of things in between.

One common thread that runs across all these designs beyond the fact that it involves the transport of electrons is the Rivian ethos of vertical integration. At Rivian, we have chosen to vertically integrate critical pieces of technology that differentiate our -- that allow us to differentiate ourselves over time. We started this journey as a start-up when we consciously chose to build our ECUs in-house. Last year, at Investor Day, we shared how this journey helped us to get to an in-house developed zonal architecture far ahead of other OEMs. Today, I'm here to talk to you about our Autonomy Hardware System, which is similarly very vertically integrated.

As RJ shared earlier today, we will be launching our Gen 3 Autonomy System late next year on the R2 vehicle platform. The hardware enabling it focuses on 3 main areas of leadership, sensors, compute and the overall product integration. Well, let's start with our sensors.

At Rivian, we have a multimodal sensor strategy that provides a rich and diverse set of data for our AI models to operate on. On the screen behind me, you can see the feed from all our sensors on an R2 vehicle. On an R2 platform, much like the R1 before, we have 11 cameras providing a total of 65 megapixels of data. That is 10 megapixels over and above what we had in R1. The cameras provide an extremely rich set of 2-dimensional data and help us see the world around us.

But cameras alone have some shortcomings. They do not perform well under non-ideal lighting conditions. This could be low light, excessive light and fog. And so, much like the R1 platform, again, we still carry 5 radars: one front-facing imaging radar and 4 corner radars. By using radio frequencies, radars are able to see in total darkness while also providing the depth and velocity of objects in their path.

Our corner radars on R2 are further improved. They support a dual mode of operation, short range and long range. In short-range mode, they have very high spatial resolution, which helps us delete the ultrasonic sensors in R2. That's right. We add sensors, but we also delete them when it makes sense. And now for the first time in R2, we're adding a third sensor, the LiDAR.

The LiDAR is an optical sensor, but its strength comes from the fact that unlike the camera, the LiDAR has an active light source, enabling it to see it much better in the dark. Another advantage of the LiDAR is that it can provide a 3-dimensional view of the world unlike cameras, which we know provide a 2-dimensional view, requiring the AI models to infer the depth, which they do, but with a lot less accuracy. So in summary, camera is the main workhorse of our sensor suite, generating the bulk of the data fed to the models, but the radar and LiDAR are critical to addressing the edge cases, which would otherwise create the long tail of problem cases.

So why do we choose to use -- introduce LiDAR now? It turns out that there are 3 main factors that make this the right moment to incorporate LiDAR. They are cost, resolution and size. About 10 years ago, LiDARs used to cost in the tens of thousands of dollars. Today, you can get a very good LiDAR for several hundreds of dollars. The resolution of LiDARs has similarly improved tremendously. The picture on the left here is the data product of a LiDAR from 2016, whereas the picture on the right, it shows the data actually from our R2 LiDAR, the one we use in our vehicle. As you can see, the R2 LiDAR data is much, much richer. Today's automotive LiDARs have point cloud densities of the order of 5 million points per second, which is about 25x better than what we could get 10 years ago.

Finally, today's LiDARs are no longer the large mechanical spinning beasts of the past. The LiDAR of today is more compact and more easily integrated into a vehicle. So let us take a look at the LiDAR integration in an R2.

There you go. This is one of our prototype R2 vehicles. From afar, it looks the same as the R2 many of you have seen and come to love. But if you zoom in closer and look upfront, you can see the LiDAR. What you see is a seamless integration and no signs of the unsightly taxicab-style bump or the TRO structure that is more commonly associated with LiDAR integrations. Our studio and design teams work very closely with the supplier to shape the face of the LiDAR in such a way so that it blends in beautifully with the R2. And now if you zoom out and look at the vehicle from the side, voila!, you'd not even know it had a LiDAR. And by the way, this LiDAR integration is camera safe. It will not burn your phone cameras, so don't worry about it.

So let's move on to compute. Before we get too deep, though, it's important to address why we chose to build in-house silicon. It's a nontrivial development effort. Those who have been involved or observed chip development efforts would know that. It's time-consuming and it requires a world-class team. The reason for doing it, though, ties back to the same reasons for building our own in-house ECUs, which is velocity, performance and cost.

With our in-house silicon development, we're able to start our software development almost a year ahead of what we can do with supplier silicon. We actually had software running on our in-house hardware prototyping platform well ahead of getting first silicon. Our hardware and software teams are actually co-located and they're able to develop at a rapid pace that is just simply not possible with supplier silicon. All of this means we're able to get to market sooner with the most cutting-edge AI product.

Secondly, we understand our application and our vehicle architecture thoroughly and are able to optimize our silicon for our use case. Note, we don't just design for today's use case, but we design with headroom for the models of the future. By building purpose-built silicon, we do not carry the overhead that comes from leveraging a design that was built for some other task and repurposed for autonomous driving. We built the silicon, so we would do a really good job at autonomous driving and physical AI problems. All of this enables us to get the best performance per dollar spent.

Finally, it's all about cost. When we design in-house, we're able to get the best cost point and the best power points. The cost reductions from our design come from the fact that this is optimized for our use case, not just the chip use case, but we look at the whole vehicle use case and as well as the fact that there's a meaningful reduction in supplier margins. Now join me as we go look for the silicon inside the R2 vehicle.

Our Gen 3 autonomy computer is the next step in our vertical integration journey and features our very own Rivian design custom silicon. It is a highly integrated solution. As you can see, there is very little on the board beyond the 2 instances of Rivian silicon, power supply and passives. The hardware and software on this computer are fully designed and developed by Rivian. This computer achieves 4x the peak performance of our Gen 2 computer while improving power efficiency by a factor of 2.5.

So let's go take a look at the chip. The Rivian Autonomy Processor or RAP1, as we call it, is the first in a series of silicon built for physical AI. It is actually much more than one piece of silicon. It's a multichip module or MCM that integrates Rivian silicon and memory die. Our custom Rivian silicon is produced on a leading-edge TSMC 5-nanometer automotive process. The star of the die, of course, is a Rivian design neural engine, which is capable of 800 sparse in 8 TOPS. TOPS stands for 1 trillion operations per second and is a common measure used to assess AI performance.

The chip was also designed with the intent of providing different cost and performance points. We can put multiple RAP processors together in a system, and they can talk to each other via custom high-speed link we call RivLink. So this gets fun. Let's take the lid off RAP1 and look under it. What you see is the RAP1 SoC in the middle, surrounded by 3 memory die spread across 2 sides. This allows for 3 independent LPDDR5 channels, but more importantly, allows for very tight integration between the SoC and memory, enabling a very clean data eye between them, which in turn then enables high memory bandwidth.

With RAP1, we're one of the first to introduce multichip module packaging for high-compute applications in automotive. This is not to be confused with systems in packages or SIPs, which have existed in automotive for a very long time. So it is well known that memory bandwidth is key for AI applications, and this tight coupling enables us to achieve a net bandwidth of 205 gigabytes per second. The MCM design also enables us to significantly simplify the design of the PCB. The PCB that it sits on no longer has to accommodate DRAM chips with critical timing constraints, which means it can be smaller, simpler and implemented with fewer layers, all of which results in a meaningful cost reduction. In summary, using an MCM style package enables us to achieve higher bandwidth and lower cost.

The SoC itself is designed to solve the needs of autonomous driving looking ahead. As mentioned earlier, the star of the show is a Rivian design neural engine capable of 800 intake TOPS. In addition to the neural engine, the SoC has a plethora of other blocks that are required to complete the system. The Application Processor Complex is implemented using 14 power-efficient Arm Cortex-A720AE cores. This allows us to leverage the rich open source software ecosystem that comes along with Arm processors. We will be the first OEM to introduce the Armv9 compute platform for automotive using the Cortex-A720AE in production vehicles.

In addition to an application processor complex, an SoC for autonomous driving applications needs a high availability safety island and compute that is real-time capable. Our SoC implements both a safety island and a real-time processing unit, which are built using 8 Arm Cortex-R52 cores. And finally, we have all the other pieces you would need for sensor processing, including an image signal processor, encoder, GPUs, et cetera.

The Rivian neural engine itself is designed to implement state-of-the-art deep learning models for perception control and planning. It is flexible and supports multiple -- supports mixed precision data formats. The neural engine has native support and hardware for the latest and greatest in AI model technology for inference. An inference chip today must support transformers and support them really well, and that is what we do. We also include support for all types of attention such as multi-headed attention, deformable attention and more. But we don't just implement transformers, the hardware has a host of other capabilities.

Some examples include support for nonlinear functions such as softmax. And of course, we do simple networks such as CNNs. We also have special hooks in the hardware to support LiDAR and radar processing, both of which can be very unstructured unlike transformers and CMMs. Earlier, I talked about designing an MCMM package with a view to optimizing for memory access. Well, we didn't just stop there. We looked at memory very holistically. Our neural engine, for example, supports weight decompression as another way to relieve the pressure on memory bandwidth. And we don't just enable running one model at a time, we can support the concurrent execution up to 4 models at any given time.

All of this would be irrelevant if we didn't have the software and tools to actually exploit the hardware. We made a significant investment in the development of tools and a middleware stack that can exploit the power of RAP1. The entire software stack is fully developed in-house. Our in-house tools include an in-house compiler that can take standard models and generate code targeting a neural engine. We also support profiling tools that can help users optimize their code.

And finally, we have an in-house middleware stack that enables us to write application code that is target agnostic. We use that same middleware stack on our Gen 2 hardware platform, and we will be using it again on our Gen 3 platform. So what makes silicon for physical AI different from general silicon targeting inference is the importance of functional safety. RAP1 was designed from the get-go to factor in functional safety in every block of the design.

We adhere to the ISO 26262 scheme for risk classification, which is also called Automotive Safety and Integrity Levels, which is a mouthful, which is why most people call it ASIL. Every block is designed an appropriate ASIL level based on the scheme and then hardware and software are implemented to ensure that the level is achieved. Even our chip-to-chip interconnect, the RivLink is protected using the scheme.

So in some cases, it means that you actually put an extra redundant hardware in the chip, which is -- does the same function twice and you cross-check the results. In some other cases, it means you put ECC on memories instead of parity, which a lot of other chips will do. So we have really a lot of hardware hooks and bells and whistles to make sure that this chip is really functionally safe.

In addition, we also have software -- it doesn't stop with the hardware, we also have software that runs on the chip when the chip is actually working in the vehicle. We have software that runs at [ KION ] to make sure the chip is still functionally safe and it runs periodically to complete the whole solution. The RAP1 chip is not meant to be just one incantation, it's -- one instantiation, it's really designed to be scalable.

While the first instantiation is a 2-chip solution targeting Autonomy in the R2 vehicle platform, it can be easily extended to solve other physical AI problems such as in robotics. It can scale down to a single chip solution for low cost or scale up to multiple chips for more performance. RivLink was specifically designed to allow multiple RAP chips to talk to each other via a high-bandwidth, low-latency interface. At data rates of up to 128 gigabits per second, RivLink allows sensor data from one SoC to be seamlessly shared with other SoCs. The scalability doesn't just end there. RAP1 was designed to also be flexible in configuration. While the system to be deployed in R2 is liquid cooled, we have demonstrated that this can be configured as an air cooled system.

I'm happy to share that we have successfully demonstrated that our silicon is robust and meeting the performance goals we set out at the start of the project. While peak TOPS are useful to indicate the capability of the hardware, a more useful measure is perhaps the ability of the system to process sensor data. We have shown that our third-gen Autonomy Hardware System is capable of processing 5 billion pixels per second of sensor data.

We are very proud to be at the leading edge of multi-modality sensing and to be continuing our trajectory of vertical integration with our RAP1 chip and Gen 3 autonomy computer. We expect that at launch in late 2026, this will be the most powerful combination of sensors and inference compute in consumer vehicles in North America.

We are now actively testing the silicon in systems and vehicles. For those of you attending this event in person, you'll get a chance to see some of our subsystem test box that test the entire hardware-software configuration. We have also integrated the hardware into our R2 vehicles and are continuing to test it extensively. I will now hand it over to James, who will show you how his team is continuing to improve the Autonomy experience for our customers and how he plans to harness the power of RAP1 to make Autonomy better. Over to James.

Okay. Thanks, Vidya. So Vidya just discussed all the amazing autonomy sensors and compute that we'll have on R2. And now I'll go into detail on some of the software that runs on them and powers the Rivian Autonomy platform. Firstly, our large driving model is trained end-to-end for millions of miles of driving sequences collected across our fleets. That's directly from pixels, radar returns and LiDAR points to trajectories. This large driving model uses state-of-the-art techniques based on transformers, autoaggressive prediction and reinforcement learning that turbocharges our velocity by allowing us to leverage innovation from the world of large language models.

It's also built entirely in-house, and this gives us unprecedented flexibility in being able to change all parts of the stack. So that doesn't -- that means we don't need to coordinate with other Tier 1s and Tier 2s to make changes. Consequently, our features improve with every update.

Finally, and most importantly, the Autonomy platform is built on a data flywheel where growth in vehicle fleets and feature adoption drives improvements in autonomy that compound over time. So let's look under the hood now and discuss the data flywheel in more detail.

We'll start with a multimodal onboard model that runs on every customer's vehicle. And the goal for our onboard sensing stack isn't just human level, it's super human level, and multiple modalities enable that, allowing our vehicle to see way beyond what a person can. By end-to-end training, the sensor data is early fused into a singular world model, a system where the sensors complement each other. They don't fight against each other.

Just like being able to hear can make you a better driver, multiple sensors can make Rivian autonomy better, enabling enhanced precision and more confident predictions. And with more sensors comes richer and better fidelity data. More sensing modalities allows us to achieve the same level of accuracy as a unimodal system, but with much less data or to surpass the unimodal system with the same amount of data. So it's a very efficient approach. Let's see how this works in practice by visualizing the output of early fusion.

So of course, we start with the cameras. Now as Vidya mentioned, our cameras are really good. Some of the highest combined megapixel counts and dynamic ranges of any vehicle on sale today. And what you can see here is that when cameras clearly sense things, the system works very well. In fact, cameras alone can handle most autonomy tasks most of the time. But for full autonomy, most of the time isn't enough. Autonomy needs to work all of the time, on a moonless night, in the snow, and here in the fog. And in those cases, cameras alone don't cut it. If we can't sense something, we can't expect the system to handle it.

And so we add radar to the mix. This is where we are with every Gen 2 R1 vehicle today. You can see that we're now able to detect much more. The system can detect more occluded objects, can assign better depth to objects and is better at estimating object velocities. The confidencies and redundancies are also much greater than with a camera-only system. But to unlock full autonomy, we need to go further. Adding LiDAR creates the ultimate sensing combination. It gives the most comprehensive 3D model of the space the vehicle is traveling through.

And as Vidya showed previously, the combination of all 3 sensors identifies more objects and can detect things more quickly. This trinity of modalities enables Autonomy features that RJ was mentioning, such as eyes-off driving and personal L4 by increasing perception quality, safety and redundancy.

Now our onboard network is designed from the ground up to flexibly and incorporate new modalities and new sensors. And together with our in-house silicon team, we're co-designing this network to run optimally on our chip. So over here, you can see where the sensors and modalities are fed into the model. Each pixel, each radar return, each LiDAR return is encoded, projected and then combined into a geometric feature space shown here in the center. And this is where the fusion happens, optimally and automatically learn through end-to-end training. There's no extra complexity added. There's no handwritten rules that need to adjudicate. The network has figured out the best way to combine this information.

And so what we're visualizing here are the semantic projections of these fused feature maps. And although this space is very abstract, humans were never really designed to see it, you can already see elements here such as other agents and lane lines showing up. So now this whole fuse sensor is fed into our transformer-based decoders to produce the well model. From the same input, the network is trained to generate all of these different outputs for the world around the vehicle.

So that includes other objects and agents shown here on the top left. It also includes dense 3D occupancy to handle short-range maneuvers and narrow negotiation. And here's where we estimate the local map around the vehicle. Note that this is quite different from typical robotaxi efforts where the map is pre-generated offline and then localized online. Those maps are expensive to generate and hard to maintain. But just like a human driver can navigate a road they've never seen before, our local map is produced and accumulated directly from what the vehicle perceives.

And in the bottom right here, finally, we have estimated trajectories. These represent the model's best estimate of how to proceed through the scene. This output will be a key technology enabler for point-to-point driving, and I'll discuss more on this piece later.

So now we're moving on to the next stop in our flywheel, the Autonomy Data Recorder. This is the system that turns real-world driving into data. Essentially, our entire Gen 2 fleet becomes a huge queryable dynamic database of driving scenarios. The data recorder runs trigger code that can far off any set of signals seen by the world model. Jaywalking pedestrians, red light runners, large animals in the road, all of these cases are examples of cases which you can mine with triggers.

We can also run more general triggers such as finding divergences between the human-driven trajectory and our predicted trajectories. These might indicate areas, for example, where the model could be improved. We can even push new triggers live to our fleets outside of the usual OTA cycle. This allows us to capture the data we need on demand with minimal turnaround time, greatly speeding up development. That's a huge driver of developer productivity. As soon as an engineer wants to find more scenarios of an event, they can mine for them immediately. And because ADR is so selective, it's also very efficient. The vast majority of boring driving data is never captured, never uploaded and never trained on.

Now once the trigger fires, all the sensor data before and after the event is captured, tagged, compressed and uploaded. And that data is then immediately available to engineers. These scenarios can be used for model training for evaluation or for replay through our simulator. Here, you can see examples of scenario clusters found automatically by our ADR system. We have examples of environmental conditions such as dusk at night. We have examples of map-based scenarios such as tight turns and mergers and also agent-based situations like animals, bike racks and semi trucks.

And after upload, all of these curative scenarios are immediately available on the Rivian Cloud. So as the feet expands and adoption increases, the size of our knowledge base is growing and compounding. From next year, this growth will be obviously further accelerated with the additional volume of R2. Now all of this data is stored securely, and we don't associate any sequences with your VIN, and if you select it, your home or your place of work. And through ADR, every sequence is already tagged without further processing.

Finally, because we store all sensor modalities, the data is incredibly rich and complete. This allows us to auto label most sequences using large off-line models, which will be too slow to run on board. In fact, the vast majority of our training data today is auto labeled. That's massively more efficient than using human annotators.

So now let's talk about ground truth fleets. You may have seen other autonomy providers, ground truth vehicles. They all typically have multiple nonproduction sensors such as LiDARs and other things strapped to the roofs and sides of the vehicles. And these fleets are incredibly valuable for training perception systems. The LiDAR data is so crisp. It's essentially used as a ground truth that is used to train the other production sensor sets. But because they're prototypes, these fleets are typically small, numbering the 10 to 100s, represented by the dots on this slide. In contrast, for the LiDAR-equipped R2, every vehicle will become a ground truthing vehicle. That's orders of magnitude more data than other OEMs. It's an incredible force multiplier for better and richer training data that massively accelerates our progress.

So let's see how we use all of this data to benefit our customers. This happens in the Large Driving Model. The Large Driving Model or LDM is an end-to-end model from sensors to driving actions. And it's based on many of the same technologies used in large language models. So LDM uses neural net transformers for processing, just like with LLMs. And the large driving model uses tokens for training, also just like LLMs. But instead of thinking about these tokens as words, there are actually small parts of trajectories that are jointly predicted and assembled together. And the large driving model also uses reinforcement learning, just like state-of-the-art large language models.

But here, instead of aligning the output with human values and intentions, we align the large driving model output for safe, performance and smooth driving. Because LDM is such a close cousin to an LLM, we can reap all of the advancements, investments and innovations being made in improving generative AI and can apply them directly to our driving task. This makes the LDM incredibly cost efficient to develop.

So let's do a deep dive into how LDM is trained by reinforcement learning. Over here, you can see the sensor data from a scenario. Here, we're approaching a stop sign. And that sensor data is fed into our transformer base encoders behind me. Then we sample multiple trajectories from this model, token by token and trajectory by trajectory. And the different tokens shape the trajectories in different ways. Once we've sampled all these trajectories, we then need to rank them. Now the one on the left here is actually the most human-like. The vehicle slowed almost to a stop, but then rolled it. A lot of our Rivian data is kind of like that. The one in the middle here, they're clearly stopping too soon. And the one on the right is just right. We're stopping behind the line and correctly following the road rules.

So here, you can see 3 different trajectories that have navigated this stop intersection. Now what we're able to do is apply our road rule rankers that can then say the third here is the optimal trajectory and then we reorder them. And then through back propagation, the model is trained to produce more of these types of trajectories in similar scenarios in the future. Now that's obviously highly simplified and you're just looking at one scenario here. But imagine this process running millions of times a second across millions of scenarios with a whole database of road rule costs and losses. And that's how LDM is trained.

And we can then distill this model into one that we can run on board. All of this work results in new models, continuous enhancements, refinements and new features that we continuously deliver to our customers. But how do we know we can release? We built a cloud-based simulator that runs the whole autonomy stack through millions of miles of real-world scenarios on every release. And that allows us to measure safety, comfort and performance in a statistically significant way without having to manually drive those miles. We also have a capability we call apprentice mode.

And before we release features, we can launch them in the background of a previous release. We can then monitor the performance of that new version compared to the human-driven miles, but also compared to the previous version of autonomy. That allows us to do an even bigger evaluation in the tens of millions of miles. And so through simulation and apprentice mode, we can rapidly build the confidence we need to ship new features and enhancements to customers.

Because the system is developed entirely in-house, we can update any part of the stack from the lowest level camera drivers all the way to the highest level motion planning code. That means that the whole stack is always improving with every release, and we have a feature road map that stretches to the highest levels of autonomy.

Now when we surveyed customers this year on the autonomy capabilities they wanted the most, the answer was resounding. They wanted more road coverage for hands-free highway assist. Now previously, we supported 135,000 miles of divided highways. But as RJ mentioned, our map today is about to grow.

Universal Hands-Free unlocks over 3.5 million miles of hands-free driving on roads across the U.S. and Canada. If there's a painted line and it's clearly marked, you can now drive hands-free. Universal Hands-Free will be part of our paid tier bundled into one simple package, Autonomy+. It's a onetime fee or you can pay month-to-month. And Autonomy+ features will be available to all Gen 2 customers for free until March next year.

And this is just the beginning for Autonomy+. We have many exciting features on the way, such as point-to-point, automatic parking and enabled by the LiDAR in R2 Eyes-Off. As our fleet continues to grow, as our adoption continues to increase, our data flywheel will continue to grow. And we've been thinking about this as a circle, but in fact, the system is better on every orbit. So a better analogy is an upward Helix, continually improving and compounding on itself.

And with that, I'd like to hand over to Wassym to discuss some of the other improvements being made in AI here at Rivian.

Thank you, James. We have made significant progress in our AI-enabled autonomy stack. But as James said, it doesn't stop here. AI runs through the core of everything we do. It's a profound platform shift, which changes our product and everything we do at the company from the way we design, we develop, we manufacture and we service our cars. This is all made possible by the Rivian Unified Intelligence, a common AI foundation that understands our products, our operations as one continuous system and personalizes the experience for our customers. So how does it work?

We revamped our vehicle operating system to be AI ready. We developed an in-house multi-agent, multi-LLM, multimodal intelligence platform. The platform is built on a robust data governance framework with security and privacy as main tenets. We have a suite of specialized agents. Every Rivian system from manufacturing, diagnostics, EVR planning, navigation becomes an intelligent node through MCP. And the beauty here is we can integrate third-party agents. And this is completely redefining how apps in the future will integrate in our cars.

We orchestrate multiple foundation models in real time, choosing the right model for each task. And we support memory and context, allowing us to offer advanced levels of personalized experience. And the architecture is natively multimodal using audio, vision and text through the same unified layer. The beauty of our architecture is the seamless integration between the cloud and the edge.

Edge AI with an embedded small language model allows us to achieve higher levels of performance, lower latency and the best conversational experience. And wait until the R2 comes. R2 will have close to 100 TOPS, Edge AI fully dedicated to in-cabin experience. This will allow us to move most of the intelligence workloads from the cloud to the edge, powering an in-cabin AI experience fully available when the car is offline.

The Rivian Unified Intelligence is the connective tissue that runs through the very heart of Rivian's digital ecosystem. This platform enables targeted agentic solutions that drive value across our entire operation and our entire vehicle life cycle. Let's start with the factory. Our diagnostics agent is the ultimate example of unified intelligence in action. It instantly connects real-time telemetry from vehicles on the assembly line and allows us to validate quality at production, identifying changes needed before the vehicle leaves the factory.

The same unified intelligence is fundamentally helping us redefine service. Because the platform is grounded in real vehicle data, every Rivian technician is now being empowered by AI systems trained on live software data, service manuals, repair histories, electrical diagrams and supply chain logistics. This is accelerating service repair time by hours, helping us to dramatically improve technician efficiency. And it doesn't stop with our technicians, the same platform that will enable in the future our customers to self- troubleshoot and resolve minor issues directly from the Rivian mobile app and from your Rivian car.

Now let me tell you more about the in-vehicle experience. This architecture will fundamentally reshape how we all interact with our Rivians. Today, I'm very excited to announce the Rivian Assistant. As you see from the beautiful wave behind me that our UX team has designed, the Rivian Assistant is fully integrated into the user experience and our in-car operating system. It's designed to understand you, your vehicle and the context you're in. And the Rivian Assistant will be available for all Rivian Gen 2 and Gen 1 customers in early 2026.

Now rather than tell you about all its features, we'll take some risk here, and then we'll do a live demo from the beautiful R1S Borealis, which is over there. So Oscar will be our main driver today. Oscar is the product manager behind the Assistant. Oscar, are we ready?

Yes, I'm ready. Thank you, Wassym. Hello, everyone. Let me show you what the Rivian Assistant can do. To initiate the Rivian Assistant, you can either hold the left steering wheel button or just say, "Hey, Rivian."

The Assistant is not just an integration of a chatbot on top of the vehicle UI. We built an agentic framework, allowing us to integrate into the larger ecosystem and bring your own digital context to the car. We started with Google Calendar. We had a ton of fun collaborating with the Google team on this project. Oscar, can we see that in action?

What's on my calendar today? What's on my calendar today?

You have 2 events today, call with Tim from 2:00 p.m. to 2:30 p.m. and meet up with Wassym from 3:00 p.m. to 4 p.m.

The agentic integration allows us to not only connect with your calendar and we did, but also take actions and manage it. Oscar, can you show us that in action?

Can you move my call with Tim to 5:00 p.m.?

I've moved your call with Tim to 5:00 p.m.

As you see, the same action would have probably taken me multiple taps and clicks and swipes to get to the same result. The Assistant can help you control your calendar in a much safer and easier way. And Google Calendar is just the beginning. The platform will expand to many more applications and many more third-party agents as they become available.

Beyond the agentic integration, our AI platform can blend your personal context into your vehicle context. In this case, the Google calendar agent is connected to other vehicle applications. Oscar, can we see a live example?

Let's go to my meet up with Wassym.

Navigating to Ferry Building. You will drive for 55 minutes and arrive around 10:50 a.m.

And the AI integration goes much deeper than navigation. Because this is an AI native operating system, it can connect with all vehicle apps, including our customers' favorites, EVR planner or drive modes. Oscar, can we see that in action?

How much battery will I have when I get to my destination?

You will have 67% battery remaining, which is about 231 miles of range.

Can you switch to a more efficient drive mode?