Welcome, everyone. Welcome to the investor conference call and audio webcast for the company's 2025 annual results.

Before we start the call, we would like to remind you that this call may include forward-looking statements, which are underlined by a number of risks and uncertainties and may not be realized in the future for various reasons. Information about general market conditions comes from a variety of sources outside of Xiaomi.

This presentation also contains some unaudited non-IFRS financial measures, which should be considered in addition to, but not as a substitute for, the company's financials prepared in accordance with IFRS. We have William Lu, Partner and President of Xiaomi Corporation, and Mr.

Alain Lam, Vice President and CFO of the corporation, to talk to us. Mr.

Lu will share recent strategic and business updates. Thereafter, Mr.

Lam will review the company's financial performance of 2025, and then after that, we will have the Q&A. Mr.

Lu, please.

Good evening. Thank you very much for coming to the 2025 full-year call.

Now, this evening, I'll be talking about three points. Review our main achievements in 2025.

Second, share our breakthroughs in hardcore tech, in particular in AI and embodied intelligence. Thirdly, we'll look ahead for the strategic direction focus in 2026.

In 2025, some of our outstanding achievements. In the year, our Xiaomi Group maintained high growth with both annual revenue and net profit reaching all-time highs.

Total group revenue reached 457.3 billion, surpassing the 400 billion RMB mark for the first time, up 25% year-on-year. Adjusted net profit reached 39.2 billion, up 44% year-on-year.

By segments, first of all, smartphones. According to Omdia, in 2025, our global smartphone shipments ranked top three.

Market share was 13.3%, remaining global top three for 22 consecutive quarters. In Latin America and Southeast Asia, our shipment ranking rose to second; in Europe and Africa, third.

According to third-party data, in mainland China, our smartphone sales ranking rose to second. On premium models, in 2025, premium models in mainland China accounted for 27.1% of total smartphone sales, up 3.8 percentage points.

In 6,000 RMB to 10,000 RMB price bracket, our market share rose by 2.3 percentage points year-on-year. We solidified our high-end base domestically and are making continuous breakthroughs in the 6 to 8k RMB ultra-high-end market.

By end of February 2026, we launched our first [Leitz?] phone for global markets priced as €1,999, a new milestone in our overseas premium strategy.

We'll continue achieving top-tier pricing in mature international markets and elevate our premium overseas sales to new heights. For IoT business, in 2025, revenue surpassed 120 billion for the first time, reaching 123.2 billion and 18.3% year-on-year growth, hitting all-time high both domestically and internationally.

For the home appliances revenue, it reached 23% plus growth with record shipments. Wearables ranked first globally, TWS earphones ranked second, tablet shipments grew 25.2% year-on-year, ranked fifth.

Our AI glasses, released in June '25, ranked third globally and first in China. We continue to drive full-category premiumization at home and abroad with overseas high-end products achieving stellar performance.

On premium strategy, in '25, our tech home appliances entered the European market, already covering Spain, France, Germany, Italy, and more. For autos, Xiaomi Auto delivered 410,000 units in 2025, far exceeding the 300,000 units target set at the year's start.

February 13th, '26, cumulative deliveries surpassed 600,000, and we are fully committed to delivering 550,000 units. In March of 19th—March 19th, we officially launched the new-generation SU7.

It features major upgrades inside and out, including across the electric powertrain, chassis, electronics architecture, etc. Within 34 minutes of launch, locked orders for SU7 exceeded 15,000, surpassing 30,000 orders after three days.

For this year's MWC, we also showcased our Vision GT concept car. Not just a concept car, this is Xiaomi Auto's latest exploration of design innovation built on hardcore technology.

We are the first Chinese brand invited to participate, and it represents recognition by the world's top simulation driving platforms. For China's auto industry, it shows that in terms of design and innovation, Chinese automakers can already compete on par with the world's best.

For hardcore tech, in '25, our R&D investment exceeded 33 billion. For '26, we plan to invest over 40 billion, with more than 16 billion for AI, embodied intelligence, and other innovation fields.

These investments build our solid product capability defenses. For AI, in 2025, we achieved breakthrough progress.

For AI, it is the era of truly useful AI, undergoing historical leap from usable to truly useful, a paradigm shift from single-task to complex-task processing, from passive to active planning, from tool attributes to ecosystem attributes. For us, with rich terminal products and use cases across smartphones, cars, home appliances, IoT, real data, value of AI far surpasses single-category companies.

Two, our foundation large models entered the leading global open-source tier. In March of the year, we released three models: Xiaomi MiMO-V2-Pro, MiMO-V2-Omni, and MiMO-V2-TTS, completing our technical foundation for the agent era.

MiMO-V2-Pro surpasses 1 trillion parameters, supporting 1 million token context windows, and ranking eighth globally in Artificial Analysis Large Model Intelligence Index, fifth by global brand. During closed beta and public launch, these models ranked first in weekly calls and held first place for many days on OpenRouter, single days as much as double the second place.

We will keep upgrading our foundation models as we move toward general intelligence. Three, we are poised to lead AI in the physical world.

Deep integration of AI and the physical world is the next frontier. We control over 1 billion hardware access points for its ecosystem.

In March '26, our phone AI agent, Xiaomi MiClaw, entered limited testing. We're the first OEM to attempt deploying [OpenClaw?]

on phone terminals, exploring the delivery of true AI phone and ecosystem to users. For auto, the new SU7 is equipped with our XLA cognitive large model, achieving mall parking navigation, complex scene understanding, and voice control, improving both driving and experience.

For SU7 Ultra, it's equipped with Super Xiao AI, a smart cockpit and advanced AI. For the home, in November '25, we launched our [Milico?]

smart home solution, giving smart homes eyes and brains and hands and feet for the first time, a pioneering real-world application of Xiaomi MiMO, laying a new vision for next-generation smart homes. Fourthly, AI is now Xiaomi core innovation engine.

In '25, with our tech forward, about two-thirds of the winning projects used AI, reimagining work across fundamental materials, chipsets, and OS, intelligent driving, tech home appliances, and more. Backed by China's strong AI industry, the coming decade belongs to China.

Also, Xiaomi is well-positioned in person, car, home ecosystem. We'll invest 60 billion RMB in AI over the next three years.

In this era, we are confident that we'll blaze a new trail for Chinese AI. For embodied robotics, it is the ultimate integration platform for AI chips, OS, and manufacturing capabilities, a high-barrier field.

In 2026, we launched a haptic-driven, precision-grasping fine-tuning model, core tech for robotic dexterous hands. Shortly after, we open-sourced the Xiaomi Robotics VLA large model, Xiaomi Robotics-0, achieving several new SOTA results.

In March, Xiaomi embodied robot began internship in our car factory, achieving 90.2% dual-side simultaneous installation success for self-tapping nut workstations, meeting production line cycle times as quick as 76 seconds with three hours of auto operation. These are just starts.

Over the next five years, we believe large numbers of embodied robots will work in Xiaomi factories. Robots will break boundary between virtual and physical worlds, but there are challenges such as cost increase, etc.

In the short term, there may be some pressure on our business, but on the other side, we will be steadfast in our strategy so that we will continue to have breakthroughs in AI chips, OS, and embodied intelligence. We're committed to scaling our global business model and advancing Chinese tech worldwide.

This person, vehicle, home ecosystem is not just a product combination, but a platform for understanding users' full-scenario data. We'll firmly seize the opportunities of AI era, and we're filled with endless possibilities and imagination.

Thank you, President Lu. As shared by Mr.

Lu, in 2025, we have achieved historical leaps. Our total revenue reached a record high of 457.3 billion, setting a company best.

This year, we achieved a year-on-year increase of 25%. Revenue in the fourth quarter hit a new single-quarter record.

Overall gross profit margin was 22.3%, up 1.3% year-on-year, historical high also. The second half of 2025 was more challenging for the first.

For the full year, our Smartphone × AIoT segment revenue was 351.2 billion RMB, up 5.4% year-on-year, which is also a new annual high. The Smartphone × AIoT segment gross profit margin also reached a record 21.7%, up 0.5% year-on-year.

For smartphones, for the year, revenue was 160—sorry, 186.4 billion RMB, accounting for 40.8% of total revenue. In '25, our global shipments reached 165 million units.

Our high-end strategy had significant results with continued product strength enhancement. Third-party data shows that in 2025, our high-end smartphone sales in the mainland of China accounted for 27.1% of our total smartphone sales, up 3.8% year-on-year.

In that, RMB 4,000 to 6,000 price segment, our market share reached 17.3%, up 0.5% year-on-year. For RMB 6,000 to 10,000 segment, our market share was 4.5%, nearly doubling year-on-year.

According to third party, in 2025, our global smartphone shipment volume ranked top three with market share 13.3%, maintaining a top three position globally for five consecutive years. In 58 countries and regions, our smartphone shipments ranked top three, and in 70 countries and regions, we ranked top five.

Despite rising memory prices in 2025, our smartphone gross margin remained relatively healthy at 10.9% for the year. For IoT, for this year, our revenue and gross margin both performed remarkably.

2025, revenue from IoT grew rapidly by 18.3% year-on-year to 123.2 billion, a new record. Both for domestic and international sales, it was at all-time highs thanks to product structure optimization and improved product strength.

For gross profit margin of IoT, it was a record high of 23.1%, up 2.8% year-on-year. By category, large smart home appliances performed exceptionally well, with revenue up 23.1% year-on-year, a record high.

Our wearable devices maintained rapid growth and an industry-leading position. Our wearable bands ranked first in global shipments, and TWS earphones ranked second.

Tablet products continue to grow fast, ranking fifth globally with shipments up 25.2% year-on-year. Internet services: we continue to expand our user base.

In December '25, our global MAU reached 750 million, up 7.4% year-on-year. Of these, mainland China MAU reached 190 million, up 10.1% year-on-year.

In 2025, our internet service revenue hit a record 37.4 billion RMB, up 9.7% year-on-year. Of this, just in the fourth quarter 2025 alone, internet service revenue reached 9.9 billion RMB.

Throughout 2025, our internet gross margin remained relatively stable at 76.5%. Advertising continued to drive internet business growth, with annual revenue of 28.5 billion RMB, a record high.

Overseas internet service revenue grew by 15.2% to a record 12.6 billion RMB, accounting for 33.8% of total internet service revenue. Next, on EVs.

Our EV and AI and innovation business segment annual revenue for this segment reached 106.1 billion RMB, surpassing 100 billion RMB in less than two years, up over 200% year-on-year, accounting for 23.2% of the group's total revenue. Of this, smart EV sales revenue was 103.3 RMB billion, with other related revenue at 2.8 billion RMB.

For the year, gross profit margin for smart EV and AI innovation business segment was 24.3%, up 5.8% year-on-year. In 2025, this segment achieved positive annual operating profit for the first time, recording an operating profit of 0.9 billion RMB.

We delivered a total of 411,082 new vehicles in 2025. In the fourth quarter alone, 145,115 new vehicles were delivered, a single-quarter record high.

The average post-tax unit price for the year was 251,171 RMB, up 7% year-on-year. Next, over the past five years, our cumulative R&D expenditure was 105.5 billion RMB, up 37.8%.

'25 alone, R&D expenditure was 33.1 billion RMB, up year-on-year. We estimate, starting from 2026, our cumulative R&D expenditure over the next five years will exceed 200 billion.

For net profit, in '25, the group's adjusted net profit was 39.2 billion RMB, a record high and up 43.8% year-on-year. For CapEx, for '25, our CapEx reached 18.2 billion RMB, up 73% year-on-year.

Of this, smart EV and AI innovation accounted for 66% and over. We continue to enhance our shareholder value, and we actively repurchase shares on the open market.

In '25, our share repurchase was approximately 6.3 billion Hong Kong dollars. Since early '26, our share repurchase totaled about 4.7 billion Hong Kong dollars.

In January '26, we announced an automatic share repurchase plan with a cap of 2.5 billion Hong Kong dollars, demonstrating our confidence in our company's long-term future. We actively practice sustainable development.

In low-carbon initiatives, in 2025, our group purchased more than 40 million kilowatt-hours of green electricity, over 10 times of last year. In '25, our PV electricity usage in our auto plant exceeded 13 million kilowatt-hours, reducing carbon emissions by nearly 10,000 tonnes annually.

On ESG ratings, in '25, for CDP Climate Change and Water Security survey, Xiaomi received a management-level B score. Also, in March '26, we achieved our best-ever score of 81 in EcoVadis gold medal, marking another recognition of our ESG efforts.

We will continue to follow a robust and enterprising operating strategy and look forward to an even greater achievement for 2026. Thank you so much.

This ends my report for this evening. Next, we can have the Q&A session.

Greetings, both. Congratulations for 2025 revenue and profits have reached new record highs.

I have two questions. One on memory.

We see that memory prices have been rising significantly. Investors what's most concerned about this for the segment of smartphones.

We notice that there are already some smartphone selling price hikes in order to offset this memory prices hike. Your smartphone sets are still the same in terms of pricing.

Perhaps in supply chain management and in inventory, you are better than your competitors so that in this challenging situation, you exceed your competitors in performance. Perhaps Mr.

Lu can explain to us for your new thoughts concerning smartphones and what are your responses to this year's challenges?

Right. For memory, yes, for each quarter, you have always been caring and concerning about this.

In different occasions, I have also talked about my views on this. In the past, I have said that this is a period that we are going through, and we have to look at 2027, and there may be high price hikes, and we have done our work in this regard.

On the other hand, my own feeling is that the cycle of hikes may be longer than I had expected. There is the AI-led demand.

Also, for memory, the cost hikes magnitude, I think, is going to be higher and much higher than I had expected. My expectation in the industry was already forward, but I think it's going to be even higher in terms of price hikes for this.

It is longer cycles, and the price hikes is going to be higher than I had expected. For all the consumer items, it is going to impact greatly, not only for smartphones, which you are more concerned about, but for some smaller-capacity categories with smaller memories units, there is a situation where there is a cut in supply even.

This is an actual situation that's facing us in reality. The impact of that, we have a calculation, which is very simple.

That is we look at the memory and its part of the product. The more it is as a part of this product category, the more it will be impacted; the less, of course, it will be less impacted.

In our categories, smartphones, tablets, notebooks, they are more in terms of proportion. On the other hand, there are some, for example, high-end smartphones, relatively speaking, it is less for memory part.

For a company, if the products will have more memory as a part of their products, then, of course, the impact will be higher; less will be less impact. This is a very simplistic way of calculation and looking at it.

In the past two weeks, looking at the memory price hikes, we already see some competitors raising their mid-price smartphones prices, and I fully understand that. I think for any smartphone manufacturer, if they do not unload to the consumer, it is very difficult to sustain this kind of price hikes.

I think it is inevitable for this. For Xiaomi, our pressure is very, very large indeed.

As I say, we will try our very best to digest this, to protect the consumer. When we can do this no more, we will have to hike our smartphones prices.

We hope that our consumers and our customers will understand this. Yes, we are slower in price hikes, but that doesn't mean that we are immune from it.

There are some competitive advantages for Xiaomi, which I can tell you. For example, in home appliances, for this category, it will be less impact.

For smart cars, EVs, it would be higher because the memory part of that is higher. On the other hand, compared to the proportion in smartphones, it will be lower.

With our variety of product segments, this is how I see it. Through these, I hope we will be able to better resolve this problem.

For smartphones, tablets, and notebooks, we're globally leading, and also in EVs as well in the past few years. For the memory suppliers around the world, we have built a very good relationship, mutual trust, and we have long-term supply contracts.

At this point, I do not feel that we have any risk of non-supply or stopping of supply. This is our competitive advantage compared to our competitors.

A third point is that, in my previous expectation, I was more pessimistic about memory price hikes, and therefore, I had been making more aggressive preparations. In that case, our inventory sufficiency was higher.

Overall speaking, for our terminal products cost, it is highly impactful. This short-term pressure is definitely in existence.

It is there.

Thank you, Mr. Lu.

This is very clear. My second question is about our vehicles.

For the new generation of Xiaomi vehicles, there had been a successful announcement. In the investor interaction, I notice that some investors feel that Xiaomi no longer talk about or announce the data, but only for the unit data.

This is more negative. For Xiaomi, what is the sale?

I think it's going to be stable, and it will be positive, and this is how I see it. Can you talk about these two investors' points of view?

Andy, let me just answer that question. For the announcement of sales from the users' side, we see some phenomena.

The first phenomenon is that for the first three days of sales, we already have achieved a significant sales: 34 minutes, 150—locked-in sales, and after three days, it was over 30,000 units sales. We have lived up to our commitment, and that is for the delivery starting from the fourth day of launch, we have already started delivery because we have already made preparations for the manufacturing of cars.

Also with some of the problems of the previous batch, they had to wait for a long time and before we could deliver to our users, those who have locked in their purchases. We have absorbed our experience, and we have had a new iteration.

As you have mentioned, why do we only disclose our locked-in contracts? We think that this is more fair.

It is not about pre-ordering or big ordering, it is locked orders. Locked orders are solid, and that is the buyers are going to take delivery of these vehicles.

This governs our manufacturing cycle as well. We think that locked contracts of purchase is a fairer way of looking at it.

That is why we made the change. In the industry, I'm sure people have their different practices, but this is what we maintain is the right way.

This is the first point. Also, concerning the locked-in contracts, let me share some information.

For the first point, it is that a lot of investors have asked about the locked orders buyers. Are they from our previous buyers or new buyers?

Most of them, we know, are new buyers. For the locked orders, they come from new buyers.

It is not the original owners who are changing to two-year-old changing from their two-year-old cars. For iPhone, about 50%, it is first generation.

For 60% of the new buyers, they are iPhone users. For our locked contracts, the progress is faster than our first-generation users.

Compared to the previous numbers, these locked orders are bigger. Also, you are very complimentary of our allocation.

About 60% of them are using the paid options. Compared to the previous generation, we think that the penetration is better.

For example, female buyers, iPhone users penetration, and also choosing different options of colors, for example, all these have been performing better in penetration than the previous generation. Thank you for your question.

Thank you, both, for picking my questions. I have two questions concerning AI.

In the past two years, there are some models and including the foundation models. For AI capabilities in the ecosystem, what is our capability?

Also for MiClaw and also for the IoT, how is MiClaw positioning, and how do we see its dance with IoT, and how is MiClaw going to be significant? Also, I would like to know how we consider AI in its users and also developers and internal to Xiaomi and its commercialization.

In LM, what are some KPI considerations for your team? You once said that in the next three years, there's going to be a 60 billion expenditure, and how is this on our OpEx and CapEx, the allocation, please?

For 2023, we have used a lot of energy to consider our AI strategy going forward. We have based it, first of all, the infrastructure for AI, algorithm, etc.

After the infrastructure, we had an exponential growth in terms of its application. Last year, we had already said that '26 with an explosive use of AI era.

From virtual, AI has moved into the physical world, and this is true to my earlier prediction. With this prediction coming true, we will be developing our AI.

LLM, we started investing in '24, and in 2025, in way more large language models, we have made a lot of progress. We have been very clear last year in saying that this year, '26, will be the year of explosive use of AI.

Last year, we were more considering about AI agent. In individual terminal, how can AI have a bigger usage so that people are able to do things that they were not able to use before?

With OpenClaw, it's allowed us to very quickly roll out MiClaw in testing. At present, from the responses of our testing, it is very, very positive.

Going forward, there is a huge market, and that is AI going into the physical world. It will be in driving, in robots, in humanoids, etc.

In this area, Xiaomi has already made the deployment, and at the end, I think it will have to serve our entire ecosystem of Xiaomi. This is a direction.

This is the large direction, and we will continue to strategically follow that. Also, you mentioned MiClaw.

For MiClaw, this AI agent of Xiaomi, this is our own developed, and it is an agent, and it is going to test our modeling capability and also our limitation, and also it's going to test how we are able to deeply integrate and also our data capability. At Xiaomi, we will do our own integration with our own data from our users.

We will also be using cloud-based data, and also there will have to be certain integration into the system. Also safety, we will be good in protecting safety.

For Xiaomi MiClaw, it is going to be the prototype for the future AI agent. Now, this is still early on.

We do not have some specific commercialization model, and I don't have any KPI for the team yet because only when it is mature will the team have certain KPI. As for the 60 billion that you talked about, yes, my colleague will answer that question.

That's right. For the 16 billion and three years 60 billion, it includes R&D and also CapEx.

Of course, in R&D, it also includes the distribution from previous years' R&D. It is our present period R&D, and also our previous R&D, and also CapEx.

There'll be three parts. If you look at the 2026 16 billion, most of it is R&D expenses, the present period.

It is 70% present period R&D. If it is CapEx plus the previous years, it would be the rest.

For the next three years, every year our CapEx will have some previous CapEx which had been detained into this year. It may be lower than the 70%.

It is for this period as well as the apportion from the previous years' CapEx or amortization from previous years' CapEx. Thank you.

Both, good evening. I have two questions.

First question concerning our IoT business. We see 2025 IoT performance have been very, very good from revenue and also in progress and advancement.

For this year, there are two concern points. They are positive.

For home appliances, new highs reached, but some investors are concerned that with the domestic situation, what do you say, the economy ramping down a bit? How do you see this?

Also, what about IoT going overseas? What is the planning?

Secondly, for vehicles, in 2025 overall, for EVs, it is already profitable. Also for future planning, for the entire year, how do you see profits for the entire year for vehicles?

I'll be talking about IoT, and then Alain will answer the other question. For IoT, for this business, because we have a number of different categories, are we talking about China market and overseas market?

For China market, I think there is an opportunity, and that is premiumization of IoT. For IoT business, for us, even though it is very large scale, but our ASP is low.

Last year, we had major progress made, but it is still far from our goal. Our watches, for example, our hair dryer, etc., I think still the average price is relatively low.

In R&D, with our investment in that, I think this year we are going high-end. I think there will be a lot, a lot of positive progress.

I think this is a huge opportunity. For major home appliances, for our washing machines, our fridge, and it is 4 points, and for air-cons, it is 10 points.

For ASP, I think there is still room for rising. For our IoT business overall, last year it is already at a very high level of IoT, as we will continue to do so.

As for overseas business, I think there is huge space for development because our market has always been in China. If going to the North American market, it's going to be three times of our original market size.

That is to see it's going to be six times if we reach our potential. It's going to be six times, or at least a few times our domestic China market.

There is a lot of empty room for us to fill in terms of overseas markets. We will send people, we will send our products, and with our IoT development, there had been a lot of overseas information and testing, and there is huge room for overseas growth and development.

For our high-ASP products, this is a high area of growth potential. This is our overseas plan.

Last year, it was 4.5 shops, and it's going to grow to 10 or 100,000—10,000 shops. I was in London, in Xiaomi Home, I was there to observe that most of the products were high-end product selling.

You think that actually, our categories are very full in range, and these overseas markets, for example, in the UK, they are selling at high end. I think there is huge room for development in the overseas market.

For the vehicle question, last year in 2025, we have delivered over 410,000 units, far exceeding the 300,000-unit target we set at the year's start. At the year's start for this will be 550,000 units for this year's delivery target for 2026.

So '26 compared to '25, there is growth. For the overall situation in '26, there is pressure.

For our expectations, etc., we are still very confident that we'll be able to reach our target. As for our target profits, you would know for this category, it is AI and new business segment, so that would include our AI investment and also new development areas.

You cannot just look at this segment as just an auto segment. It includes other new businesses in this segment.

At present, the new businesses are still in the investment stage. As I've mentioned, in AI, for example, this year we will continue to increase our investment in AI, including in robotics.

We will increase our investment there as well. You have to look at two areas: one, auto in this segment, and secondly, AI plus new business investment.

For this particular segment, last year it performed very well. For this year, as auto growth and also in other areas, the kind of fruit that we're going to pick from them, the results, this is going to be an encouraging segment.

Thank you very much for the answers. Next question, please.

Thank you. Can you hear me fine?

Yes, we can. Thank you.

I have two questions. I would like to know, for the Middle Eastern situation recently, I don't know whether for the overseas business, including IoT, handsets, has it impacted you?

Are there certain logistics and also cost of raw materials had presented issues to you? Also, another question concerning your gross profit margin: for this year, for handsets or smartphones, is there a principle, let's say, under what kind of profit level will you be keeping it or making certain choices for adjusting the price?

For smartphone handsets to protect your gross profit margin? Also, for vehicles, there may be some pressure, as you have mentioned before.

How do you think this will be making your performance better than your competitors? Are there certain safeguards for profits?

Also, AI, IoT, because premiumization is a strategy of yours, you did say that for this kind of premiumization, what is your plan for AIoT, please?

For Middle Eastern conflict, it's nothing that we want to see, certainly. I hope that there will be a solution for it because it's going to impact industries and economy around the world.

It's going to impact big ways. As for Xiaomi business, the Middle Eastern overall situation, for revenue from that part, it is only in the single digit as a contribution to our profit.

In terms of the market from the Middle East, it is a small proportion of our overall at Xiaomi. It is controllable from that regard.

At the same time, we also see that the oil situation from the Middle East, we already see some plastic pellets, and other raw materials is influenced or impacted. Overall speaking, it is still controllable.

That's for your first question. Second question, concerning cars, IoT, and also smartphone handsets, the gross margin, and what is our pricing strategy?

For this, I would say, first of all, for memory. For memory component, if we are to quantify that, I would say in the past for a quarter, for example, we expect it to be at a certain high price, but actually it's going to be at an even higher price, rising even higher.

It's very hard for us to quantify that. Even for a small increase, because it is such a big part of the product, it will be impacting the cost of the product a lot.

It is very difficult to foretell early on like what we have been able to do before. For smartphones, for this category, I think given our size and also our market share, it is very important for us.

If you say whether there's a principle, it would be that we try our best to make some balances.

My consideration is that we want to maintain our market position. That is very important to us.

Because for smartphones, there are very few listed companies in this category, I'm not being able to get a lot of accurate data from the industry competitors. From my years of experience, I know the following.

For example, in vehicles, their absolute number is, as a part of the overall car price, it is less in terms of its impact, but not as big as a smartphone part. For IoT, it is also even lower in terms of internal memory price impact.

Different categories will have different impact. What is big impact would be smartphone, notebook, and also tablets; less so for smart cars, less so for vehicles, that is; and even less so for IoT.

Thank you, both, for this opportunity to ask a question. I have two questions.

On AI business, MiClaw has been introduced, and I was fortunate to have joined the launch. That was excellent.

Can you give us more guidance, and that is at what point of maturity would you in big model or MiClaw to commercialize them, and how would that be like in revenue? That's the first question.

As for the second question, for this year, for chip, for example, do you think that you can have some progress to update us?

For MiClaw, I have already talked a lot about it. I'm sure, as a user, you will have felt this product.

You were there at the launch, and it's given us a lot of nice surprises, but it's a new product. We still have a lot of there's a lot of room for improvement.

I myself also have gotten a lot of responses, we have to improve on it. I think the iterations will be very fast, and it will be a new version in a few days in a few days.

It's going to be it's going to have high-speed iterations. For Xiaomi, overall speaking, for AI, it's still going to gonna be contribute to our users.

For the commercialization of AI, I would say, at this point, it is still too early, even though our large model efficiency is high, our token commercialization, for instance. From absolute numbers, it may be a little bit high.

At this point today, I would say commercialization is too early to talk about for us. Also, you already see our XLA model.

This is the XLA model, which is a cognitive large model. SU7 is equipped with it.

From at present, internally, from our testing, it is very, very good in performance. For these two models of cars, I am a user.

I personally would use if I use them, and if I have any questions in terms of use, I will put these cases to the team. I personally experiences auto driving functions.

It will be going forward step-by-step. Also our auto driving, whether it be models or chips, we already have some deployment.

Integration terminal to terminal, and when we have this ready, it's going to give a lot of new experience to our users. I don't know whether you've been driving our cars.

Yes, watch out for its progress as we integrate more and more of our models into them for automotive.

Thank you for the opportunity. A lot of questions have already been asked by others.

I have two questions to ask concerning new business investment and also IoT. For new business, with the models iterations faster and you said that there'll be more investment into AI, for expenditure on new business, is chips side will be adjusted, or do we see that there will be new chips introduced?

Second question on IoT: just now it's been mentioned, for tablet and notebook, there will be some cost impact. Will there be something like smartphones pricing strategy, and that is to emphasize the experience for users, please?

For this year, we have increased a lot of R&D expenses. For chip, it is a long-term strategic capability of ours.

I have already said that this is a platform capability, chip, because it's going to provide capability for a lot of product types and product categories. Even though in AI, we have increased our investment, but we have not slackened our investment in chips.

Actually, many of the chips, they are part of our big AI strategy. We will definitely be steadfast with it.

As for PC and tablets, we will follow more or less the same strategy as for smartphones. Please also notice that for the notebook that we have just launched, the Xiaomi notebook, after four years of development, it is selling very well.

The demand is even higher than we had expected. When we launched this notebook, we already knew that the memory part will be increasing in price, but on the other hand, the response has been so encouraging.

Because if the product strength is good, even though it is more expensive, the users will be able to adopt it, and they'll accept it. I think product innovation, technological capability, these are important.

Even though memory is hiking in prices, but still we will have ways to make our products attractive. In 2026, through our efforts, I think our company and our management team will still be able to deliver good performance for everyone.

Thank you. We end the meeting here.

Thank you very much for your participation. Thank you for your support for the company, Xiaomi.