Li Auto Inc. (LI) Earnings Call Transcript & Summary

Thanks, operator. Hello, and welcome — good evening to those joining us now and good morning to those elsewhere. This is Li Auto's First Quarter 2026 Earnings Conference Call. Earlier today, a press release went out containing the company's financial and operating results, and that same release was also made available on the company's IR website. On this call, our Chairman and CEO, Mr. Xiang Li, along with our CFO, Mr. Johnny Tie Li, will open things up with prepared remarks. For the Q&A portion, our President, Mr. Donghui Ma, and our CTO, Mr. Yan Xie, will be joining as well. Before we move ahead, we'd like to remind everyone that forward-looking statements will be part of today's discussion, made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995. Inherent risks and uncertainties come with forward-looking statements. Because of that, what the company actually achieves could differ materially from the views expressed today. Additional details about risks and uncertainties can be found in certain company filings with the U.S. Securities and Exchange Commission and the Stock Exchange of Hong Kong Limited. Unless applicable law requires otherwise, the company takes on no obligation to update any forward-looking statements. Also, please be aware that discussions of unaudited GAAP financial information, as well as unaudited non-GAAP financial measures, are included in Li Auto's earnings press release and on this conference call. On the IR section of our website, you can find Li Auto's disclosure documents, which include a reconciliation of the unaudited non-GAAP measures to comparable GAAP measures. Our CEO will begin his remarks in Chinese, and an English translation will follow once he has finished all of his remarks. With that, I'll now hand the call over to our CEO, Mr. Xiang Li. Please go ahead.

Hello, everyone. Li Xiang here. I appreciate all of you for being on today's call about our earnings. During this year's first quarter, our deliveries shifted onto a path of growth. Between January and April, Li Auto climbed back to the number one spot in sales across Chinese brands as well as in China's new energy vehicle segment for vehicles priced at RMB 200,000 and above. For our BEV model, the Li i6, monthly sales have settled at 20,000 units each month, putting it in the top 3 across all BEV SUVs. We introduced the all-new Li L9 on May 15, with deliveries beginning May 17. There are 2 versions of the all-new L9, Livis and Ultra, priced at RMB 509,800 and RMB 459,800, respectively. Our main objective with this all-new generation Li L9 is to secure the market position of a flagship SUV — key elements of how a flagship product is perceived, including styling, suspension and chassis, range extender and electric powertrain, plus intelligence and computing power, set the benchmark for what the next generation of flagship SUVs should have. In a mere 2 weeks, the L9 Livis locked in more than 10,000 orders at transaction prices above RMB 500,000, and we anticipate holding a market share above 20% in the RMB 500,000 and above NEV SUV segment. Come June, our communication and promotion push will center on the Li L9 Ultra, with the goal of seizing a 20% market share in the RMB 400,000 to 500,000 NEV SUV segment. The all-new Li L9 represents the start of a wave of new product introductions for the Li L Series. Later in June, we'll roll out the all-new Li L8, a remarkable 5-seater flagship SUV. As the 5-year version of the all-new Li L9, it's a total overhaul versus the prior generation, and it's no longer a step down from the L9; in our view, the all-new Li L8 might be the best-handling large activity worldwide, while offering the most comfortable 5-seat experience in its class. With the all-new Li L9's launch, we've successfully and completely rolled out our in-house Maho M100 chip alongside the MINDVLA model. Mass producing our full stack hardware software solution was a major milestone for us. We're the first company in China to deliver full functionalities on a brand-new chip in its first ever on vehicle deployment. The Maho M100 chip is the 5-nanometer automotive-grade AI inference chip built on an AI-native dynamic data flow architecture. This distinctive architecture and outstanding treating power created a long-term technological moat for us; through an integrated hardware and software design, our chip provides 3x the effect of computer power per unit cost. What's more, the Maho M100 chip lets us deploy our newest MINDVLA model on our vehicles. Parameters in this new model rose tenfold compared with the earlier version. Rolling out Maho M100 and MINDVLA is only a starting point. Looking ahead, as bigger models plus data training reach a higher position and faster frame rates, we anticipate an enormous jump in the automotive and autonomous driving experience. The event on May 15 centered mainly on hardware and vehicle performance, and in our view fully showing our progress in software and intelligence would call for a devoted 2- to 3-hour session. A separate launch event is being planned for June, dedicated to sovereign AI; we'll take the time to give an in-depth walk-through of the real-world experience spanning in-cabin interaction, salvation model, autonomous driving, system agents and our Maho chip. We look forward to delivering a deep dive into the many things software and embodied AI can bring into our lives. Please stay tuned; with the steady rollout our core technologies and our updated product portfolio, we maintain our full year sales growth target of 20%. With that, I'll turn the call over to our CFO Johnny, to walk you through our financial performance. Thank you.

Thanks for that, Li Xiang. Hello, everyone. Given how limited our time is, I'll concentrate my remarks on the quarter one financials. Unless I say otherwise, all numbers I mention are denominated in RMB. Should you want additional detail, including the U.S. dollar equivalent, please refer to the earnings press release we issued. Total revenues for Q1 were RMB 23 billion, down 11.4% versus a year ago and 20.1% versus the prior quarter. Of that amount, RMB 21.5 billion was vehicle sales, representing a 12.7% year-over-year decrease and a 21% sequential decrease. The year-over-year reduction was driven mainly by a lower average selling price resulting from a shift in product mix. The quarter-over-quarter decrease largely reflected reduced vehicle deliveries due to seasonality surrounding Chinese New Year, together with a lower average selling price resulting from a shift in product mix. Cost of sales in Q1 was RMB 21.2 billion, up 2.7% from a year earlier and down 10.4% from the prior quarter. First quarter gross profit was RMB 1.8 billion, down 66% versus a year ago and down 54.8% versus the prior quarter. New core margin in Q1 was 6.1%, versus 19.8% a year earlier and 15.8% in the prior quarter. The year-over-year and sequential decreases were driven mainly by the shift in product mix. Gross margin in Q1 was 7.9%, versus 20.5% a year earlier and 17.8% in the prior quarter. Operating expenses in Q1 totaled RMB 4.8 billion, down 4.8% year-over-year and down 13.8% quarter-over-quarter. R&D expenses in Q1 were RMB 2.7 billion, up 8.3% year-over-year and down 9.8% quarter-over-quarter. SG&A expenses in Q1 were RMB 2 billion, down 19% year-over-year and down 22.6% quarter-over-quarter. The decreases both year-over-year and sequentially were driven mainly by reduced employee compensation and lower spending on marketing and promotion. Loss from operations in Q1 was RMB 3 billion, versus RMB 271.7 million of income from operations a year earlier and RMB 442.6 million of loss from operations in the prior quarter. Operating margin in Q1 was negative 13%, versus 1% a year earlier and negative 1.5% in the prior quarter. Net loss in Q1 was RMB 2.3 billion, versus RMB 646.6 million of net income a year earlier and RMB 20.2 million of net income in the prior quarter. Diluted net loss per ADS attributable to ordinary shareholders was 2.26% in Q1, versus diluted net earnings of 0.62 a year earlier and RMB 0.01 in the prior quarter. Separately, through today we have bought back 17.5 million Class A ordinary shares in aggregate, which includes 7.3 million ADS, for total consideration of USD 148.1 million; vehicle margin was 6.1% in the first quarter, versus 19.8% in the same period last year and 16.8% in the prior quarter. Let's now turn to cash flow and the balance sheet. Operating cash flow for Q4 showed a net outflow of RMB 6.1 billion, versus RMB 1.7 billion consumed in the prior-year period and RMB 3.5 billion generated in the preform. In Q1, free cash flow came in at negative RMB 7.4 billion, against negative RMB 2.5 billion a year earlier and RMB 2.5 billion the quarter before. Our cash position remained solid, ending the quarter at RMB 94.3 billion. With such a strong tax position, we continue returning value to our — to shareholders through the USD 1 billion buyback program unveiled in March. To date, we have repurchased a total of 17.5 million Class A ordinary shares, including 7.3 million ADS, for a total consideration of USD 148.4 million. Vehicle margin in the first quarter was 6.1% versus 19.8% in the same period last year and 16.8% in the prior quarter. Now, regarding our business outlook for Q2 2026, the company expects deliveries between 95,000 and 100,000 vehicles, with quarterly total revenues projected to fall between RMB 24.1 billion and RMB 25.4 billion. This guidance reflects the company's current and preliminary view of the business landscape and market conditions, and it may change. That concludes our prepared remarks. I'll now turn the call over to the operator to kick off our Q&A session. Thank you.

My first question concerns the L9. What does order inflow look like for the auto L9? For that variant, the waiting period has now stretched to 9 to 11 weeks. Could you tell us how the company has arranged production capacity for this model? Also, within your second quarter delivery guidance, what sales mix is targeted for the L9? That wraps up my first question.

To begin with, order patterns for the L9 are quite clear. The lived version, our top seller, makes up more than 90% of total orders, while the Ultra version, already at full load, takes the remaining under 10%. That mirrors how customers recognize our newest advanced technology, their readiness to pay for features and performance, and it also demonstrates our stable footing in the market above RMB 500,000 — a very encouraging trend for the brand. Going forward, we plan to boost promotion of the Ultra version and keep refining the order mix. Next, regarding production capacity: both the all new L9 and L8 will be built at our — base, and the 2 cars can be flexibly adjusted along the production line. Over the long term, we feel confident about our ability to manufacture these 2 models. May and June will serve as the ramp-up period for these 2 cars, with monthly production capacity landing between 4,000 and 5,000 units per month. Right now, the 2 tone body color of — plus certain unique parts on this model are supply constraints, slightly supply constrained. We are working around the clock with core suppliers on solutions so these cars reach our customers as soon as possible. Meanwhile, we have ample production capacity for the Ultra version, and we can adjust production according to market demand. Last, on L9 deliveries in Q2, factoring in the production ramp-up, we expect around 8,000 units delivered between the middle of May and the end of June. Once we fully ramp up in Q3, we are confident that the all-new generation online will reach a delivery level over the previous generation outline.

< > For my follow-up, I'd like to ask about profitability. How do you see margins shaping up in Q2? And looking at the whole year, when should we anticipate a distinct turning point for earnings, and considering raw material costs have gone up, can we still get back to profitability within this year? On a separate note, the Auto i6 now makes up almost 60% of all vehicle sales — what does that mean for the overall gross profit margin level? And finally, would you mind sharing the gross margin target for the L9 plus the other models coming out later this year? That's my follow-up question. .

Tim, John here. I'll handle this one. Several things weighed on our gross margin in Q1, among them the model refresh cycle — we have to refresh starting from the line — plus a bigger share of I6 and L6 deliveries in the overall mix, and the purchasing tax subsidy on the I6. That said, once the all-new L9 launches and deliveries begin, we anticipate gross margin recovering roughly 10% in Q2. Over the full year, as our model refresh cycle wraps up and we fine-tune production ramp-up, we look for gross margin to keep improving. Our top priority this year is getting the refresh of the Li L series done successfully. It's encouraging that the Li L9 — our flagship, with its capabilities and technology leadership — is earning strong recognition in the market, and we're capturing share in roughly the half-million price range. The all-new L9 doing well gives us a firm foothold in that price segment going forward, building on what the original Li L9 achieved. This year's all-new Li L series, together with our EV lineup — including the I9 we launched featuring extensive future in-house developed technologies — establishes a solid base for us across the next 2 years. Thank you.

Let me respond in English. Regarding vehicle performance — what are the standout differentiators and the real cost savings delivered by the large model? And where does the company's autonomous driving system head next?

This is Jen, and I'll take that question. Against our ADAS 8.0 release, the 9.0 release, which runs on the M100 chip we built ourselves, delivers a clear step up. The main gains show in choices made during complicated situations, with a more human-like lightness of control, in both the longitudinal and lateral directions, plus a smoother, more comfortable ride and drive overall. Version 9.0 marks our first AD release operating on chips made in-house, and it already ranks among the top offerings in this fiercely contested market, yet this is only the start; thanks to the new platform, sensors gather data at greater accuracy and over a longer range, while the M100's strong compute lets us execute bigger and superior algorithms. So with this new platform we advance data, compute, and algorithms simultaneously; and this — that is what will produce a far quicker jump in what our autonomous driving can do. Turning to what comes next for autonomous driving: First, we'll keep expanding our input data and precision models, so that more driving-related semantic information can be fed into the newer network, since this lets the model perceive far more signals directly from the sensors. Second, we'll sharpen the model's cognitive abilities, particularly how well it learns short-term cost and effect relationships. That enables the model to move past mere behavior fitting, letting it form human-like judgments in more complicated urban traffic situations. Last, we'll push the system much further into the execution stage, with more compute latency optimizations coming from our in-house operating system and a fully drive-by-wire chassis, so the car governs motion with greater precision and reacts more quickly. This — that means the autonomous driving system will come across as more confident and, more importantly, safer. On top of that, since we develop the software and hardware jointly, our in-house M100 chip provides triple the computing power of the prior generation platform at half the cost. At a comparable cost, we get 6x higher effective computing power. With the same model, our input trend rate has tripled, and inference rates have risen by an even larger margin. Our aim is to equal the performance of Tesla's FSDB 14 in the United States during the second half of this year. The stronger AI inference system constructed around M100 chips hands us a solid foundation for achieving that. Thank you.

I'll go ahead and use English for this. From the time the store partner program was put in place, what kinds of subsidies have shown up in key metrics — for instance, average monthly sales story in the south < > area, out for employees, and expense ratio at pilot stores — relative to before the reform? Has the impact the program currently has on sales volume come in line with what was expected? Also, how does the company quantitatively assess the program's effect booked in sales for Q3 and later on?