Earnings Call
WeRide Inc. (WRD)
Earnings Call Transcript - WRD Q4 2025
Operator, Operator
Good morning, and good evening, ladies and gentlemen. Thank you for standing by, and welcome to WeRide's Fourth Quarter and Full Year 2025 Earnings Conference Call. Please note that today's event is being recorded. Please note that the Chinese interpretation is for convenience purposes only. In the case of any discrepancy, management statements in their original language will prevail. Joining us today are WeRide's Founder, Chairman and CEO, Dr. Tony Han; and CFO and Head of International, Ms. Jennifer Li. Before we continue, I'd like to refer you to the safe harbor statement in the company's earnings press release, which also applies to this call as today's call will include forward-looking statements, including WeRide's strategies and future plans. These forward-looking statements are made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995. Forward-looking statements involve inherent risks and uncertainties. The company's actual results could differ materially from those stated or implied by these forward-looking statements as a result of various important factors, and please refer to the Risk Factors section of the company's Form 20-F filed with the SEC and announcements on the website of the Hong Kong Stock Exchange for a full disclosure of these risk factors. The company does not assume any obligation to update any forward-looking statements, except as required under applicable law. Please note that all numbers stated in management's prepared remarks are in RMB terms, and we will discuss non-IFRS measures today, which are more thoroughly explained and reconciled to the most comparable measures reported in the company's earnings release and filings with the SEC and the Hong Kong Stock Exchange. The company's unaudited financial and operating results were released earlier today via Newswire and can be found on the company's IR website. And with that, I'll now turn the call over to the company's Founder, Chairman and CEO, Dr. Tony Han. Please go ahead, sir.
Tony Han, CEO
Hi, everyone. Thank you for joining us today. As a global leader in autonomous driving, we delivered strong results over the past year with a record high revenue of RMB 685 million, growing 90% year-over-year, driven by expanding robotaxi deployments as well as robust demand for robobus and other autonomous driving products. Robotaxi revenue has increased 210% year-over-year, reflecting the continued commercialization of our technology. By today, our global robotaxi fleet size has reached a new height of 1,125 vehicles. We are seeing encouraging momentum across both our domestic operations in China and our international markets. As we continue to scale, I'll walk you through the key developments driving our growth. Let's first turn to China, which continues to be one of our most important operating markets. During the period, we have advanced toward a more data-driven autonomy stack by incorporating end-to-end learning, leveraging large-scale data training and world model simulation as well as improving algorithm generalization. We also made solid progress in operating efficiency, fleet expansion, service coverage and user adoption. On the operation front, total cost of ownership decreased by approximately 38%, driven by reduction of vehicle BOM costs and improvement of operating efficiency. The BOM cost of our latest robotaxi GXR is cut by 15% enabled by the adoption of our cost-effective computing platform, HPC 3.0. At the same time, our remote assistance human-to-vehicle ratio improved from 1:10 in 2024 to 1:40, currently making labor cost marginal and further strengthening unit economics. Fleet size and service coverage also continued to improve. Our commercial and testing fleet in China has grown to more than 800 robotaxis, covering over 1,000 square kilometers across key urban areas. As fleet density increased, we also saw meaningful improvement in service performance. Average daily orders per vehicle reached 15 trips over the past 6 months, rising to 26 during peak periods. Average passenger waiting time declined to under 10 minutes, reflecting stronger demand and improved vehicle utilization. We continued expanding our presence in Beijing and Guangzhou, focusing on populated areas, connecting key transportation hubs such as major airports and railway stations, along with further penetration into downtown areas. In Guangzhou, for example, our robotaxi service is now available in Tianhe District, one of the city's most active commercial hubs. On the product side, we launched a flexible free PUDO feature, allowing riders to select any pickup or drop-off location within the service area. At the same time, we broadened access through major mobility platforms, including Amap, WeChat and Tencent Mobility. Encouragingly, we are seeing substantial increase of registered users of WeRide's robotaxi service. Take the fourth quarter as an example, the year-over-year growth of registered users exceeds 900%, reflecting stronger user acceptance and engagement. Overall, we believe WeRide's long-standing leadership in robotaxi technology, combined with extensive real-world operational experience forms a durable and hard-to-replicate competitive moat, continues to resonate with users. As our operations scale, we expect further growth in vehicle utilization and user adoption going forward. Let's now turn to our international operations, where we continue to build on our progress expanding our global footprint. Today, we have deployed autonomous vehicles in 12 countries with official permits already granted in 8 of those markets. This progress demonstrates our ability to navigate complex regulatory environments while meeting local requirements. Taken together, these milestones position us as the most globally deployed autonomous driving company today. Europe is shaping up to be another major growth area for us. Back in November 2025, we achieved a historic milestone, receiving Europe's first-ever driverless robotaxi permit for passenger service from Switzerland's Federal Roads Office. That head start gives us a real advantage as we look to expand further across Europe in the coming years. Apart from this, just this month, we added another country to our global map as we announced our entry into Slovakia. We are deploying our full product lineup there, launching the country's first-ever AV program. Slovakia is also the fourth country for our European footprint. The Middle East remains one of the most promising regions. In Abu Dhabi, we secured the world's first city-level fully driverless robotaxi commercial permit outside the U.S. in October 2025. Today, our service covers about 70% of the city's core area and passengers can book through various categories on the Uber app, including Uber Comfort, Uber X, and the new autonomous category, Uber's first dedicated autonomous ride option globally. In that quarter, we also launched commercial robotaxi rides in both Dubai and Riyadh. We've also started pilot operations in Ras Al Khaimah, giving us a presence in a third UAE Emirate. In Asia Pacific, we are building early momentum with a strong local presence. In Singapore, WeRide and Grab began autonomous vehicle testing in the Punggol district back in November 2025. Our robotaxi GXR and robobus are expected to open to the public by April 2026, making Punggol Singapore's first residential neighborhood with an autonomous shuttle service. A big part of what makes all this possible is our disciplined approach to international expansion. Our strategy focuses on markets with supportive regulatory environments and favorable economic conditions that are conducive to sustainable operations. Today, our international robotaxi fleet has surpassed 250 vehicles, and we continue to scale deployments across key international markets, including Abu Dhabi, Dubai, Riyadh, Switzerland, Slovakia and Singapore. Each new market we enter becomes a regional blueprint for expansion into adjacent markets. Looking ahead, we expect our global robotaxi fleet to reach 2,600 by the end of 2026, subject to regulatory approvals and market conditions. These milestones represent the early stage of our long-term vision to deploy tens of thousands of robotaxis worldwide by 2030. While robotaxi services are at the core of our business, we are actively diversifying into other areas of autonomous mobility to build a broader portfolio and generate additional revenue streams. Across our diversified applications, WeRide's global AV fleet spanning robotaxis, robobuses, robovans and robosweepers grew rapidly from 1,089 vehicles at the end of 2024 to 2,113 as of today, strengthening our global presence across 12 countries. Take our robobus business as an example, it has seen impressive growth in 2025, increasing by 190% year-over-year. We've entered multiple European markets, including Switzerland, France, Belgium, Spain and Slovakia, where labor shortages in public transit systems are creating real opportunities for autonomous shuttles. By leveraging our technology to meet this demand, we're able to address a pressing need while also expanding our market presence. Additionally, our L2+ advanced driver assistance system, WePilot 3.0, is being adopted by leading OEMs and Tier 1 suppliers, including Chery, GAC and Bosch. This system uses a one-stage end-to-end architecture with vision-based perception, enabling vehicles to operate safely and efficiently across diverse driving conditions. In the second China Urban Intelligent Driving competition, hosted by D1EV.COM, Chery Exeed, a mass-market passenger car model powered by WePilot 3.0 made history by winning first place in 3 cities' competition stops. This business expansion helps us capitalize on the growing demand for L2+ ADAS technologies even outside our L4 fleet. Innovation remains at the heart of our strategy. Building on our strength in L2+ end-to-end systems, WeRide is converging its technology stack toward L4 by integrating end-to-end learning with large-scale simulation and real-world data loops to enhance generalization and edge case handling. Our latest development, WeRide Genesis is a general-purpose simulation platform that integrates physical AI and generative AI. WeRide GENESIS generates high fidelity driving scenarios much more efficiently, reducing simulation time for complex edge cases from days to minutes. This boosts our development efficiency and enables us to quickly test and deploy improvements to our systems while significantly reducing on-road testing costs, reinforcing the technological moat underpinning our L2 to L4 product portfolio. On the hardware side, we've taken our partnership with Geely Farizon to the next level. The latest GXR as a purpose-built factory pre-installed autonomous vehicle is delivering higher safety consistency and lower unit cost. The upgraded model incorporates our proprietary HPC 3.0, a high-performance computing platform, which is a more compute-efficient architecture. And here's the number that really shows the efficiency gain. Per vehicle production time is now under 10 minutes. These advancements will give us a significant edge in terms of both vehicle performance and scalability of our global robotaxi fleet. With that overview, I'll now turn the call over to our CFO, who will walk you through our financial performance and outlook in more detail.
Jennifer Li, CFO
Thank you, Tony. Hello, everyone. Before we dive into the financials, I want to highlight that all figures are in RMB and comparisons are year-over-year unless stated otherwise. Now let's discuss our fourth quarter and full year 2025 financial performance. Looking back, Q4 2025 will be remembered as a defining chapter in WeRide's journey as we continue to make steady progress in scaling operations, improving unit economics and advancing our technological leadership. Our fleet and geographic coverage grew significantly, showing strong momentum as a global leader in autonomous driving. Now let me walk you through the financial performance that reflects these developments. In Q4 2025, total revenue was CNY 314 million, representing an increase of 123%. Product revenue increased 309% to CNY 211 million, mainly driven by increased sales of robotaxis and robobuses. Service revenue increased 15% to CNY 103 million. For the full year 2025, total revenue increased 90% to record high CNY 685 million. Product revenue and service revenue were CNY 360 million and CNY 325 million, respectively, representing an increase of 310% and 19%, respectively. Robotaxi continues to be one of the most crucial drivers among the business lines. In Q4 2025, WeRide has reached another record high quarterly robotaxi revenue since its foundation as its robotaxi revenue has increased 66% to CNY 51 million. For the full year 2025, robotaxi revenue also achieved an annual record high of CNY 148 million with a 210% growth year-over-year. This growth is powered by our asset-light model. We provide the autonomous driving brain while mobility and fleet partners handle operations and vehicle ownership. This keeps vehicles off our balance sheet, enables efficient scaling and aligns incentives to deliver a consistent user experience. Our robotaxi revenue is also supported by a combination of ongoing service fee and recurring ride-hailing revenue share from platform partners. Our global expansion in robotaxi operations has firmly established us as a leader in the rapidly evolving autonomous mobility landscape. In particular, our international business is becoming an increasingly important driver of group revenue. For the full year 2025, overseas markets contributed approximately 29% of total revenue, spanning a diverse set of geographies, including the Middle East, Europe and Asia Pacific. At the same time, our focus on operating efficiency has allowed us to sustain healthy margins even as we scale. In Q4 2025, group level gross profit increased 74% to CNY 89 million with a group level gross margin of 28%. For the full year 2025, group level gross profit was CNY 207 million, representing an increase of 87% with a group level gross margin of 30%. These results highlight the strength of our business model and reinforce our confidence in our global expansion strategy. In Q4 2025, operating expenses increased slightly by 2% year-over-year to CNY 655 million, with R&D representing 63% of the total. For full year 2025, operating expenses decreased by 11% to CNY 2.04 billion, with R&D representing 67% of the total. The decrease in operating expenses was primarily driven by lower administrative expenses, partially offset by increases in R&D and selling expenses. As our primary investment area, the rise in R&D spending reflects our focus on long-term technological leadership, while disciplined cost management kept overall operating expenses under control. To break down further, R&D expenses increased by 29% to CNY 411 million in Q4 2025 and increased by 26% to CNY 1.37 billion for the full year 2025, primarily driven by our continued investments in top-tier talent and the expansion of our next-generation data center infrastructure, building a high-performance backbone to support our L2+ to L4 autonomous driving capabilities. This continuous R&D commitment is essential to maintaining our competitive edge and driving future growth. Administrative expenses decreased by 29% to CNY 217 million in Q4 2025 and decreased by 48% to CNY 596 million for the full year 2025. The decrease was mainly driven by lower share-based compensation and partially offset by growing professional service fees, personnel costs and depreciation and amortization expenses. Selling expenses increased by 76% to CNY 27 million in Q4 2025 and increased by 37% to CNY 74 million for the full year 2025. These increases are in line with the growth of our business. Our net loss narrowed by 6% to CNY 5,565 million in Q4 2025 and narrowed by 34% to CNY 1.65 billion for the full year 2025. Building on our successful Hong Kong IPO in Q4 2025, we ended the year with a strong capital position. As of the 31st of December 2025, we had total capital reserves of CNY 7.13 billion, comprising CNY 6.97 billion in cash and cash equivalents and time deposits, CNY 144 million in investments in wealth management products and CNY 19 million in restricted cash. We maintained short-term bank loans of CNY 324 million to support daily operations. With disciplined cash deployment, this level of capital provides a solid operating buffer and underscores our financial strength to support continued growth. On the 23rd of March 2026, our Board of Directors authorized a share repurchase program under which we may repurchase up to USD 100 million of our Class A ordinary shares, including in the form of American depositary shares over the next 12 months. This authorization also reflects our commitment to shareholders and our confidence in the long-term value of WeRide. Looking ahead, we are moving forward with clarity and conviction. By the end of 2026, we expect to have 2,600 robotaxis globally, marking the first phase of our path toward tens of thousands of robotaxis by 2030. As our fleet scales, we see a clear path to extending our proven deployment model to more cities worldwide, supported by a strong balance sheet, relentless focus on operating efficiency and deepening partnerships, we are well-positioned to lead the next chapter for the autonomous driving industry.
Operator, Operator
We will now take our first question from the line of Kai Xiao of CICC.
Unknown Analyst, Analyst
This is Ben from CICC. So I have 2 questions. First one is quite a few OEM and Tier 1 peers are also working on L4 robotaxi such as Horizon Robotics and Momenta. So what's your view on this matter? And the second one is on the L2+. So recently, vRS ADAS showed an industry-leading capability in a highly challenging real-world competition. So could you share the key differentiation between vRS ADAS capability versus the peers?
Tony Han, CEO
Thank you very much, Kai, for these 2 important questions. And I truly appreciate you asked these 2 questions. So first of all, the first question is about quite a few OEM and Tier 1s you just mentioned, like a few names like Horizon and Momenta. They claim that they are working on robotaxis, and what's my view? So first of all, I think as a first mover and the industrial leader of autonomous driving companies like WeRide, I have to say we welcome other players to join this competition. That means the whole ecosystem and the whole industry are really attractive and there's a great market potential. So only in a very interesting and profitable market can you see competition. Having that said, I just want to remind some competitors or new players of the difference between ADAS systems, which they are very familiar with, and the L4 driverless robotaxi system, which they are not familiar with. A lot of time, people have to do some contractual extrapolation from what they have in the ADAS system and think that by just increasing the reliability and a little bit of speed of their system, they can achieve driverless operation. My answer to that kind of extrapolation is that they haven't seen the difficulties that lie ahead. Why am I in a position to say that? Because WeRide is the only company currently, to my best knowledge, that is good at both ADAS systems and driverless operations or robotaxi. For example, in the China Urban Intelligent Driving competition just finished a few days ago, we won the championship once again. WeRide is the only company that has won this championship 3 times in a row; we made history. This means that in that competition, the companies you mentioned, like Momenta and Horizon, all entered the competition, but WeRide won 3 times over them. So we know a lot about the ADAS, and we are very familiar in the ADAS area where they are comfortable. On the other hand, if you want to claim yourself to be an L4-level autonomous driving company or a robotaxi company, you have to have substantial operations. What do I mean? You should have at least 50 driverless cars running in cities with populations of at least 1 million people. If you just have 10 cars with a safety driver behind the steering wheel, you cannot claim yourself as a robotaxi company. Of course, a lot of people can claim they are working on something; that's great. But to ensure you are a real player in the field, you need to demonstrate your capability in driverless operations. Therefore, I want to emphasize that the technological barrier between L2++ and L4 is huge. To be a serious player in the robotaxi field, you need to showcase your capability in driverless operations. Besides, there are other factors such as hardware maturity, software integration, full redundancy architecture, regulatory approvals, and operational stability that traditional ADAS companies haven't experienced. My view is that WeRide, along with some other leading L4-level robotaxi companies, has a strong advantage. We welcome this kind of competition, but we are very confident about our current leading position and international market footprint. That creates a huge gap, and I don't foresee any newcomers catching up soon. This answers your first question. The second question is about our recent extraordinary results in the China Urban Intelligent Driving competition. What differentiates us from our peers? I think the ranking and scores tell all. This kind of competition is held once every month or every 2 months. Historically, only one company has won this kind of championship twice in a row; that was Huawei. At that time, they had a significant advantage. But no company has ever won this kind of competition 3 times in a row. Since November last year, December, and this March, we have won 3 times in a row. The competition has been entered by all the well-known autonomous driving companies in China, including XPeng, Li Auto, NIO, Xiaomi, Momenta, Horizon, Huawei, and Zeekr. But WeRide won this kind of competition 3 times in a row. This demonstrates our capability. What is the secret sauce behind it? We have several. First, one of our one-stage end-to-end systems trains based on both synthetic and real data from our L4-level autonomous driving robotaxi and our L2++ level data, which helps a lot. Additionally, our unique one-stage end-to-end architecture is different from others, and that gives us a big advantage. Thirdly, the data generated from our GENESIS model is superior and can produce data according to our needs while also reducing our data collection costs by 75%. With all these technologies combined, we have a leading position in the ADAS field. That concludes my answer to these 2 questions. Again, thanks for the question, Kai.
Operator, Operator
We will now take our next question from the line of Ming-Hsun Lee of Bank of America.
Ming-Hsun Lee, Analyst
I also have 2 questions for you. So first, could you please elaborate on the robotaxi expansion plan of both China and international markets? And what is the delivery schedule for the 2,000 vehicles signed with Geely Farizon? And second question, could you also share your forward-looking plan for the Middle East, given peers' market entry, geopolitical conflict and also the involvement of the partnership with Uber? Specifically, what is the timeline of the deployment of 1,200 vehicles with Uber?
Jennifer Li, CFO
Okay, I'll take the first question. I guess you can take the second question. Thank you, Ming. Both China and international markets are core to WeRide's growth strategy. As we just mentioned, we recently signed an extended agreement with Geely Farizon for an additional 2,000 upgraded robotaxis, which we call GXR, in 2026. So by the end of last year, our fleet has surpassed over 1,000 robotaxis. Taking into consideration the phased delivery of the newly pre-installed GXR and the retirement of older vehicles we have, we expect to reach around 2,600 robotaxis globally by the end of this year. China definitely is a cornerstone of our business, where we focus on cities with very supportive policies for autonomous driving and high population densities; those Tier 1 cities in China, like Guangzhou and Beijing, both fit this profile very well. In Guangzhou, for example, we plan to gradually scale towards city-level operation like we already have in Abu Dhabi. We are also expanding into another major Tier 1 city in China, and we look forward to sharing more details soon. On the international side, the Middle East continues to be a very strong base for us. We run the largest robotaxi fleet at the city level in Abu Dhabi, and we are the first to start robotaxi public operation in Dubai and in Riyadh as well. So we'll continue to maintain our strong momentum in the Middle East. Europe is also a key focus; our robotaxi obtained the first and only driverless operation permit in Europe from the Swiss government last year. Building on this momentum, we are looking to further expand into markets like Madrid and potentially one more core city in Europe this year. Of course, we'll continue to expand our robotaxi operation in Zurich as well. Tony, do you want to take the next question?
Tony Han, CEO
Sure. So just to remind everyone, the second question is really about our future plan for the Middle East and the timeline of deployment of the claimed 1,200 vehicles with Uber. Right now, we have around 200 vehicles in the Middle East. Together with Uber, we plan to add at least 1,000 more by 2027. We expect to be the first to reach 1,000 vehicle scale in this region. It's really not easy, and we have spent a ton of effort on planning and sometimes retrofit and redesign our vehicles and undertaking lots of technology efforts to meet this target. Operationally, we are carrying out business as usual. I think you have noticed we are already running a city-level full driverless robotaxi service in Abu Dhabi, and we aim to do the same in Dubai this year. We are currently the only company offering robotaxi service to the public in both cities, and that first-mover advantage is meaningful. Our scale, regulatory progress, strong partnerships, and safety record all support this position. Concerning the geopolitical tensions we are currently experiencing, we are monitoring developments closely, with a continued focus on the safety of our local teams and maintaining reliable operations. So far, we have not seen any material impact on our business. Our global footprint and diversified presence also give us confidence in navigating potential uncertainties. Regarding our partnership with Uber, our 5-year, 15-city rollout plan remains on track, and we expect to announce new city launches in due course this year.
Operator, Operator
We will take our next question from Tim Hsiao of Morgan Stanley.
Tim Hsiao, Analyst
This is Tim from Morgan Stanley. I have 2 questions. I think the first one is basically, I want to follow up on the L4 competition because at NVIDIA GDC earlier this month, we noticed that WeRide again showcased the robotaxi GXR. I think the model is powered by the NVIDIA Hyperion platform and the Thor SoC. Although we noticed the collaboration with NVIDIA has helped WeRide to effectively reduce costs and accelerate global expansion, these days, I think NVIDIA also supplies similar ready solutions to many of WeRide's competitors as well as to automakers for their long-term L4 development. So my question is, how does WeRide address the challenge of robotaxi homogenization in the long term and can keep successfully differentiating?
Tony Han, CEO
Thank you, Tim. This is a great question. So first of all, I think currently, many players define their autonomous driving systems based on NVIDIA's Thor AGX platform. However, I want to emphasize, just making announcements like adopting NVIDIA Thor AGX is easy. But creating a reliable and workable autonomous driving system based on NVIDIA's Thor AGX is very challenging. The WeRide team partnered with Lenovo and NVIDIA spent 2 years designing our HPC 3.0 autograde. We actually produced the first-of-its-kind computational platform for robotaxis with a computational power of 2,000 TOPS. You can look around and see whether anyone else has this kind of computational platform of 2,000 TOPS with redundancy; it’s not easy. We have built a lot on top of this autograde computational platform. Additionally, we have built our integrated platform GENESIS for simulation and data collection. GENESIS can generate a lot of data and be integrated into training, which can then be evaluated with HPC 3.0. So that is an extremely advanced system. In the long run, I don't believe that a general, very generic universal platform can democratize the entire industry to ensure that other less technologically advanced players can become significant players in robotaxis. This was tried before; I just want people to remember that 4 or 5 years ago, Baidu rolled out this Apollo platform. The goal was very similar to what NVIDIA wants to do: roll out some open-source code and let everyone work on it. However, to our best knowledge, no driverless robotaxi fleets have been developed using Apollo. Even Baidu has resorted to its own closed-source platform. Whether that open-source framework can be utilized by third parties to deploy reliable driverless operations is an unproven concept, and I'm quite skeptical about that. We are very confident in our competitive edge in the next 5 years and the next 10 years. I think we have a significant lead that will only get larger. That is my answer to your question, Tim.
Tim Hsiao, Analyst
Thank you, Tony, for sharing all the details. My second question is also a quick follow-up regarding our global partnership with Uber because I think most of us noticed that WeRide has formed a close strategic tie with Uber for global expansion. Yet at the same time, I think Uber has a ride-hailing platform, and the company continues to onboard more robotaxi service providers like Rivian, Zoox, Motional in the U.S. and Wayve, etc. in the rest of the world. So facing this kind of dynamic of both cooperation and competition, how does WeRide ensure its long-term share of ride-hailing orders in overseas markets can keep growing and stay a major supplier to Uber? Lastly, how does WeRide plan to enhance its capability globally?
Jennifer Li, CFO
Tony, you want to take this?
Tony Han, CEO
Yes, I'll take this question. So Tim, first of all, Uber is an important shareholder and partner for WeRide. We believe their incentive is to maximize our WeRide's robotaxi utilization. As a matter of fact, with all the partnerships they have signed to date, you can only book robotaxis from Uber on the Uber platform from WeRide and from Waymo; that's it. We have a very concrete plan to scale together. In key markets, like the Middle East, we have the first and foremost operating rights, and we are going to launch 1,200 robotaxis, which will be fully delivered by 2027. This will be a very large and definitely the largest robotaxi fleet outside China and the U.S. In addition to that, we operate robotaxis, robobuses, robovans, and other products globally. We act more like an infrastructure partner to local governments. For instance, in the Middle East, we are bringing our robotaxi, robobus, robovan, and robosweeper all together in one go. We are the infrastructure partner to those local governments, not just a supplier for Uber. We own most of the autonomous driving licenses. The government directly issues licenses to us. Also, in the presentation we just did in the video, you can see that our global taxi partners worldwide include, besides Uber, we also have Grab and TXCI and a number of local partners. We have different local partners in different local markets. So just to emphasize the licensing moat, we are the only one holding autonomous driving permits in 8 countries. Based on our knowledge, this is definitely the broadest in our industry. This creates a very high barrier to entry. Furthermore, even with Uber and other product categories like whenever we have public service deployments, we are building the WeRide brand recognition directly with end consumers. So if you talk in Singapore, Abu Dhabi, in the Middle East, and in Dubai, people recognize the WeRide brand quite well, I would say. The bottom line is that we are not just one of Uber's vendors; we are their equity-linked key operators and key robotaxi providers in key markets with proven unit economics. We believe our multi-scenario, multi-country footprint will make us an essential part of the urban mobility infrastructure, not just a replaceable supplier.
Operator, Operator
We will now take our next question from Jiajie Shen of JPMorgan.
Jiajie Shen, Analyst
Congratulations on the very strong results. My first question is regarding fleet utilization. How do you expect it to grow in China? And what are the implications for unit economics? My second question is regarding overseas business. Your global business is showing great progress, and this is truly impressive. Could you please share more about key figures of the overseas business in 2025? And what are the 2026 guidance of key operational and financial metrics?
Tony Han, CEO
Okay, I will take the first question. The first question is about the fleet utilization growth in China and the implications for unit economics. So first of all, we are taking a number of steps to improve utilization, such as raising fleet density in existing service areas, rolling out free pickup and drop-off, and extending operation hours. We are seeing that translate into results. User adoption rates have been strong. Our robotaxi user cases in China grew over 900% year-over-year in Q4. This is an extraordinary number, and I don’t expect we can keep this kind of number every year, but it is indeed an impressive statistic. Our vehicle utilization is also trending up. Currently, each vehicle averages about 15 trips per day, with peak days reaching around 26 trips during special occasions like festivals or Valentine's Day. The average trip distance is about 5 kilometers and pricing is roughly RMB 2 per kilometer, which represents a 30% to 50% discount to traditional ride-hailing. This is a deliberate promotional strategy as we scale region by region and drive our user adoption. As we expand to citywide coverage like what we are working on in Guangzhou and deepen integrations with platforms such as Amap from Alibaba, WeChat from Tencent, and Tencent Mobility, we expect pricing to move closer to the standard rate of around RMB 3 per kilometer. At this point, we see a clear and achievable path to improving unit economics in China. Over the years, as our fleet size and coverage expand, we are targeting 25 trips per vehicle per day because autonomous driving vehicles, like our robotaxi, do not require a driver. Therefore, we can keep operating as long as we can take orders. At steady state, we expect contribution margins in China to be over 40%.
Jennifer Li, CFO
The second question is regarding overseas business and some of the forecasts, right? Yes, our international business is growing at a really impressive rate. For Q4 2025, the overseas revenue was up 140% year-over-year, contributing 31% of the total revenue, with the gross margin at around 40-50%. For the full year 2025, our overseas revenue grew about 305% year-over-year, accounting for roughly 29% of total revenue, with a gross margin of almost 50% as well. The overseas market offers significantly stronger growth and profitability potential. For example, our Middle East subsidiary is already profitable on a stand-alone basis. Looking ahead to 2026, we expect the revenue to continue to grow at a healthy pace. We are on track to reach our global robotaxi fleet goal as well as targets for different regions. We believe we can achieve this goal by the end of the year. Regarding gross margins, we expect them to remain relatively stable for the international business. On the cash side, operating cash flow may increase modestly at the group level since we are continuing to invest in talent and R&D to support the long-term core growth of the company. We'll provide more detailed guidance in the upcoming earnings release. But this directionally reflects how we are thinking about building up the business and our financial forecast.
Operator, Operator
We will now take our next question from Leo You of CLSA.
Leo You, Analyst
It's Leo from CLSA. Also 2 questions from me. The first is, could you please walk us through the key cost reductions and technology innovations behind the upgraded GXR? And how would that drive further unit economics improvement going forward? And secondly, we are also very glad to see that you announced the share buyback plan. Could you please elaborate more on the thought process and how you're going to execute this share buyback plan going forward?
Jennifer Li, CFO
Okay, I'll take both questions. Thank you, Leo. The first one is on the cost reduction for GXR. I will answer this question from the total cost of ownership perspective. As we mentioned earlier, the total cost of ownership for the China fleet has declined by as much as 38% in 2025. The two main drivers are as follows. First, there’s a significant improvement in the remote assistance ratio from 1:10 in 2024 to 1:40 by the end of 2025. We have a similar efficiency gain for the grid operator, so we are observing efficiency improvements on the operational level. Secondly, we have seen an overall 30% reduction in the BOM cost, including the upgraded HPC along with the pre-installed new GXR robotaxi. We are also starting to benefit from larger volumes, which is beginning to yield real benefits. As our fleet expands, we see meaningful cost reductions through volume procurement. Additionally, on the software side, I want to emphasize that the game-changer here is really our WeRide GENESIS. This platform allows us to handle edge cases much more efficiently and accelerates the iteration cycles. It enables us to deploy in much larger operational design domains and with less remote assistance required. Together, this will lead to a very meaningful improvement in our cost structure. Regarding the share buyback, today our Board of Directors authorized a share repurchase program effectively as of March 23, 2026, which is today. We may repurchase up to USD 100 million worth of our Class A ordinary shares on both the Hong Kong Stock Exchange and NASDAQ over the next 12 months. This is subject to the scope and limit of the repurchase mandate granted by the shareholders of the company on March 13, 2026, and approval of a similar repurchase mandate to be put forward to shareholders at the upcoming 2026 Annual General Meeting. Our proposed repurchases may be made at market price or in privately negotiated transactions or block trades, depending on market conditions and in accordance with applicable rules and regulations. That answers both of your questions, Leo.
Operator, Operator
We will now take our next question from Xinyu Fang of UBS.
Xinyu Fang, Analyst
Congratulations on delivering solid revenue growth and operational expansion. My first question is about revenue structure. We noticed that there have been quarterly fluctuations in the contribution of product and service revenue. How should we think about the revenue structure of the company going forward, both in the near term and medium term? As for my second question, apart from the CNY 100 million share buyback program, could you please share a little bit more on the future cash deployment plans of WeRide?
Tony Han, CEO
Okay, I will take the first question. In 2025, our robotaxi contributed 22% of the total revenue, the robobus contributed 34%, and our L2++ ADAS and data service contributed 29%. Together, these three pillars accounted for around 85% of our business, with robovans and robosweepers making up the remaining 15%. That's the detailed numbers. But if you look forward, I can say that robotaxi, robobus, and L2++ will remain our core growth engines, supported by strong synergies across our integrated autonomous driving ecosystem. We have a platform strategy; robotaxi is the fastest-growing segment with significant scalability and improving unit economics, especially overseas. We expect its revenue share to increase over time. Just to emphasize, WeRide is inherently a robotaxi technology company. We have several products, but robotaxi is the core. The robobus is currently our most geographically deployed business and benefits from clear synergies with robotaxi in both regulation and commercialization. L2++ continues to gain traction with partners like Chery, GAC and Bosch, serving both as a revenue stream and a platform to validate our technology. Meanwhile, the robovans and robosweepers provide complementary value with lower-cost fixed-route operations, which also help us enter and educate new markets. Overall, this diversified portfolio gives us both scale and balance, which we see as a key strength. If you look at all autonomous driving companies in the world, WeRide is unique; only WeRide adopts this technology.
Jennifer Li, CFO
Okay, I'll answer the last question. For the cash deployment side, we ended 2025 with a little over CNY 1 billion in cash reserves, which is a strong foundation to support our expansion. Our net cash burn rate is less than USD 200 million based on past patterns. At the same time, our revenue growth is accelerating, as we can see from today's results, and our operating cash flow is becoming increasingly important for funding. As we scale, we are generating cash ourselves at a relatively fast speed. Overall, we are in a very solid liquidity position, and we will continue to invest with discipline while maintaining ample runway to execute our growth strategy.
Operator, Operator
Thank you. Due to time constraints, I'll conclude the call today. Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.
Tony Han, CEO
Thank you very much.
Jennifer Li, CFO
Thank you.