Open-source operating systems, Li Auto aims to become the leader of the "new world"

Author | Wang Xiaojun Editor | Zhou Zhiyu The transformation of the automotive industry is moving towards a deeper direction. Recently, Li Auto open-sourced its self-developed automotive operating system called "Li Auto Star Ring OS," becoming the world's first company to open-source an automotive operating system. By choosing to open-source this system, Li Auto aims to become a leader in the second half of the automotive industry's intelligent transformation, just as Tesla led the first half of electrification. On April 11, Li Auto's CTO, Xie Yan, stated in interviews with Wall Street Insight and others that as cars transition from Smart to Intelligent systems, there is a greater need for centralized systems, and both hardware and software need to address these issues. "In the past, there was no unified consensus, but I believe that as we move forward, some consensus may emerge, which is what we refer to as everyone seeing the same map." Li Auto's open-sourcing is aimed at becoming a key leader of this "map," defining new rules, and creating an ecosystem for the automotive era; this is a significant strategy for Li Auto. Previously, the industry judged that Li Auto's move was to become "the Android of the automotive industry." Xie Yan explained that while there are similarities between Star Ring OS and Android, it is not a user operating system like Android, and Android itself is also bound to certain user applications, not a pure operating system. Star Ring OS is a purely intelligent system centered around intelligent driving, capable of running many applications such as intelligent driving, vehicle control, and chassis, without binding Li Auto's applications. "What Star Ring OS is will be determined by what 'applications' run on it. Here, 'applications' do not refer to the applications in the internet industry, but rather applications related to the automotive field, such as intelligent driving," Xie Yan stated. This statement also clarifies the parts that Li Auto will open-source and those that will not. The part of Star Ring OS that will be open-sourced is "equivalent to a set of internal brain, cerebellum, nerves, and immune system for robots," while the "applications" running on the system, such as intelligent driving, will not be open-sourced. According to Li Auto's current vision, if the industry widely adopts this system, it will greatly improve industry efficiency. It is well-known in the industry that a few years ago, the chip shortage caused companies like Li Auto to lose billions. At that time, Li Auto was still using the AUTOSAR operating system, which held an absolute dominant position in the industry. If a shortage of MCU chips needed to be replaced, the adaptation and verification cycle of the operating system and the new chip generally required more than six months Now, Li Auto wants to improve the overall adaptability of industry systems by open-sourcing its own operating system. Regarding chips, Xie Yan stated, "When we launch the Li Auto Star Ring OS, we will collaborate with proactive chip companies whose hardware can directly adapt to the Li Auto Star Ring OS, allowing us to download the operating system directly onto a development board for use." The same applies to other software and hardware aspects. After open-sourcing, for users, copying Li Auto's ready-made operations will save a significant amount of upfront R&D costs, reduce the work of "reinventing the wheel," and significantly improve R&D efficiency, thereby enabling the industry to quickly reach a common node. Li Auto Chairman Li Xiang previously stated, "The Star Ring OS can help the automotive industry save 10 billion to 20 billion yuan in redundant R&D resources each year and break the 'black box' technical barriers of closed-source systems, reconstructing the technological monopoly of international suppliers." The vision is certainly beautiful, but what intrigues the industry even more is why Li Auto is putting so much effort into creating an operating system only to open-source it? After all, this approach to open-sourcing an operating system is not common in the entire industry. With competition so fierce, every bit of technological progress from each company becomes key to standing out in marketing. Moreover, different automakers have varying capabilities in system development. How will Li Auto's "Star Ring OS" attract other automakers and form a commercial closed loop? Li Auto's idea does seem somewhat idealistic. Xie Yan stated that now is a suitable time for this operating system to be open-sourced. In the first half of electrification, industry innovation has mostly been fragmented and distributed. Tesla has opened many patents because it believes that relying solely on one company's strength to drive the electric vehicle industry will slow down development and that more companies need to join in. Now, for the automotive industry to achieve the leap from Smart System to Intelligent System, it also requires joint efforts from the industry to move more steadily and quickly. Thus, Li Auto's choice to open-source its operating system aims to accelerate the industry's journey toward intelligence, with Li Auto wanting to be a leader in this process. If the Star Ring OS can be accepted by more suppliers and automakers in the future, the adaptability of software, hardware, and systems can also improve, forming an efficient cycle. Currently, Li Auto hopes the industry can gradually accept the Star Ring OS. Xie Yan stated, "If a Tier 1 thinks this is good and uses it entirely, that would be great, and we can spend time discussing it together. If it's only partially used, I still think it's very valuable." The following is a dialogue with Li Auto CTO Xie Yan, edited without affecting the original meaning: Q: Can you summarize "Star Ring OS" in a few simple sentences? Xie Yan: Many people will compare "Star Ring OS" to Android, but it is not Android (a user operating system); it is a set of intelligent systems that can run many applications such as intelligent driving, vehicle control, and chassis. What Star Ring OS is will be determined by what "applications" run on it, referring not to the applications in the internet industry but to applications related to the automotive field, such as intelligent driving, etc Question: Li Auto will open source its intelligent driving system after July. Many car companies are currently developing intelligent driving, and there are many third-party suppliers (for intelligent driving). What is the significance of Li Auto's open sourcing? For example, does this mean that companies do not need to develop (intelligent driving) and can directly use Li Auto's system? What impact and shock will this bring to the industry, such as challenges for third-party suppliers? Xie Yan: Intelligent driving and automotive operating systems are two different concepts. Intelligent driving itself is akin to an application on the automotive operating system, similar to an App. The intelligent driving that the industry commonly refers to mainly concerns intelligent driving algorithms, which include end-to-end, VLM, VLA, etc. Each company will develop these intelligent driving algorithms based on its business needs, and they are usually not open-sourced, possessing customized attributes. Li Auto's intelligent driving is also not open-sourced. The automotive operating system is the foundational software of a platform system that can support different intelligent driving systems from various companies. Question: Who will use this system? I believe this overall model should be developed and then packaged and sold to other competitors. If problems arise, Li Auto would be responsible, but in reality, Li Auto does not charge for it. Does this mean that if issues arise, they cannot be identified? Xie Yan: This is a very good question, and it is indeed not an easy task; no one has done this before. First, I believe that to accomplish something, you need "timing, location, and harmony." Previously, the "timing" was not right; in the last phase, everyone was fragmented and separated, without these demands. If you wanted to add a function, you could just find a place to add it. However, as cars transition from Smart to Intelligent systems, there is a greater need for centralized systems, whether in hardware or software, to solve these problems. If there are no good solutions today, everyone will solve it on their own, but now that there are better solutions, everyone can use them. There was previously no unified consensus, but I believe that as we move forward, some consensus may emerge, which is what we refer to as everyone seeing the same map. For capable competitors, being able to reference this system is also valuable. Some may think that open sourcing the system has no value for Li Auto, but in fact, it does. If we see some functions or directions for solving problems placed in the open-source code, influencing other car companies through the code, and everyone moves in the same direction, the industry will move in that direction. More people will see a consensus and gradually focus. We do not expect the open source to be fully utilized on the first day, although that is our ultimate goal. If a Tier 1 company finds this system good and uses it entirely, that would also be great, and we welcome that; we can also spend time discussing it together. If it is only partially used, I still think it is very valuable; it is a blueprint. Question: What considerations does "Li Auto Star Ring OS" have at the business model level? Have you considered the return rate? Xie Yan: We do not hope to gain returns directly from open sourcing. We neither sell automotive operating systems nor interfere with how users use them. We only hope that everyone joins in, uses freely, contributes freely, and moves forward together with Li Auto on the same path, so that we can move fast enough to ultimately realize the vision of creating an Intelligent system The business model of Li Auto's Xinghuan OS is different from Google's, as Google is a service provider that requires Android manufacturers to incorporate its services after open-sourcing, thereby gaining commercial returns. However, Li Auto is not like this; we only hope to achieve our vision faster and better together with suppliers and peers. Reporter: Is there a specific model? Do you plan to provide services? Xie Yan: No, we do not gain commercial returns directly from open-sourcing. I believe open-sourcing is just a starting point, an Intelligent System. Whether we call it a car or a space robot, to move forward, a lot of software and hardware collaboration is needed. Therefore, we hope to move forward together through open-sourcing, which requires a lot of collaboration from many people, allowing everyone to see a picture, a blueprint for the future. Android actually achieved this in the smartphone sector, but in the automotive field, how to move forward specifically and what functions we need to provide are still somewhat unclear. I believe with Li Auto's Xinghuan OS, everyone can see the future direction of progress on this "map" or this set of protocols. Once everyone reaches a consensus on the prospects, they can invest in research and development without hesitation, without worrying that hardware or software investments will go unused, which is a significant value of open-source operating systems. Question: What potential benefits does open-sourcing bring to Li Auto? Xie Yan: To move towards the future, some things need further iteration from us, and some require collaboration from everyone, meaning the industry needs to reach a consensus. Tesla once opened many patents because it believed that relying solely on one company's strength to drive the electric vehicle industry would slow down development and that more companies needed to join in. Just like without Android, the industry's development would be much slower. Each company developing its own system with different rules would lead to a lot of redundant construction, with various expressions, operations, and UI (user interface) methods differing, which would further trouble users. From this perspective, a true open-source project is like a map or a public protocol, where everyone can continuously move forward on this map, reducing a lot of redundant construction in the industry, improving efficiency, and moving more steadily and quickly. For Li Auto, with the accelerated industry development, we still have confidence in leading the way. Question: Li Xiang mentioned at the Zhongguancun Forum that Xinghuan OS will help the automotive industry save 10 to 20 billion in redundant investments each year. How is this calculated? How can it provide other manufacturers with an attractive cost reduction expectation? Xie Yan: Different manufacturers have different perspectives and needs. Our operating system has multiple advantages: first, in the face of some common technical challenges in the industry, some manufacturers often need to invest hundreds of people and take many years to solve these issues, which occupies a large amount of their R&D expenses and resources. Secondly, our system has undergone overall optimization and can now achieve cost control at the level of billions, mainly because we are reducing the work of "reinventing the wheel," which can save a lot of R&D costs and significantly improve R&D efficiency In addition, open source can reduce the consumption of R&D resources, as there are more compatible hardware options, providing greater flexibility and potentially lowering costs from this perspective. Therefore, cost optimization is a comprehensive result, not just due to a single point. Q: Which fields and roles does Li Auto hope to see join? What are the top three prioritized suppliers? For example, suppliers that work on electronic and electrical architecture? Xie Yan: We welcome all automakers and suppliers to come together to use and contribute. There are some common issues in the industry that we need to solve, such as real-time performance and safety. The more people use the Li Auto Star Ring OS, the lower the costs, so there is no ranking, and we even welcome companies from non-automotive sectors to use it. Q: No ranking, is there a priority? Xie Yan: There is no priority. The Li Auto Star Ring OS is a comprehensive and extensive automotive operating system that involves general computing, AI computing, Ethernet, hardware abstraction layers, and includes multimedia, cameras, etc. It does not mean that we need all modules to be complete; users can flexibly choose based on their needs and still achieve good functionality. Q: How many companies are currently developing their own automotive operating systems? After Li Auto announced open source, have other companies reached out, and what is the industry's acceptance level? Xie Yan: Currently, the industry either uses supplier solutions or develops their own, so there are a few companies developing their own automotive operating systems. After we announced open source, there has been a high level of attention in the industry, but no details have been disclosed, and many people want to know the specifics. As we release the white paper and code, once everyone sees the actual system and documentation, they will truly start to develop and use it, and then we will know the industry's acceptance level. At the same time, people are also concerned about the open-source method and whether using the Li Auto Star Ring OS will lead to control issues. We will use a permissive Apache License (Note: a free software license published by the Apache Software Foundation). As industry contributors, Li Auto will open source the code and technical documentation, will not charge fees through open source, will not interfere with how the code is used, and will not control user data. DeepSeek is a great example; they open up weights, and users can deploy services without any fees and without interference or control over users, and we will do the same. Q: If an automaker integrates the Li Auto Star Ring OS into their platform and models, what do they need to do? Can Li Auto provide some specific guidance? Xie Yan: This is what we need to do well. First, we will have a white paper and technical documentation; second, when we launch the Li Auto Star Ring OS, we will collaborate with proactive chip companies whose hardware can directly adapt to the Li Auto Star Ring OS, allowing us to directly download the operating system onto a specific development board for use. Q: Will Li Auto have a customer service team to provide support? Xie Yan: We will not have a customer service team; we operate on a community model. We do not take anyone's job; Tier 1 companies can choose to use the Li Auto Star Ring OS to build their own systems and serve their customers We will not charge fees for the open-source Li Xiang Star Ring OS. Q: Existing open-source operating systems are not widely used in the industry. What do you think is the reason? How does Li Xiang avoid this problem? Xie Yan: Currently, there is no complete automotive operating system in the industry like Li Xiang Star Ring OS that covers vehicle control and intelligent driving. Q: Other industries also have open-source operating systems, but they are not widely used. Xie Yan: Because the systems from other industries may not be suitable for smart cars; directly using them does not necessarily mean they are useful. Q: What are the reasons for their ineffectiveness? Xie Yan: There are two reasons. First, the system itself does not solve core problems; some open-source projects merely reinvent the wheel without addressing many issues or creating value. Second, the system may be outdated and does not address future problems. While solving our own issues, we also see some future challenges. Open-sourcing the operating system is not a one-time deal; open-sourcing is just the starting point. We hope to work with the industry to solve many problems together and move faster and further. Q: When did Li Xiang decide to open-source the operating system? Earlier, you mentioned that Li Xiang realized this system is complex and that it is difficult for one company to do it well. Did you think this from day one, or did you encounter particularly difficult problems along the way that made you realize the need for industry collaboration? Xie Yan: Let me clarify that Li Xiang Star Ring OS is not ineffective; it is currently performing well. It is useful, and being useful is the premise for open-sourcing Li Xiang Star Ring OS; otherwise, we wouldn't open-source it and embarrass ourselves. We see more opportunities in the future, from Smart systems to Intelligent systems. With the development of AI, we have greater space and demand for transformation. This is our purpose for open-sourcing and moving towards the future together with suppliers, car manufacturers, and companies from other fields. Xie Yan: From day one of developing Li Xiang Star Ring OS, we have been discussing whether we could open-source it if we did well. For engineers, good work is something they hope to gain more recognition for; of course, this is a technical perspective and unrelated to business. This early rough idea was just talk, but later we increasingly understood what the next stage of automotive operating systems should look like and felt that the industry needed such an operating system. Therefore, we open-sourced Li Xiang Star Ring OS, hoping to gain recognition from everyone and help the industry see and achieve the vision for the next stage together. Q: Open-sourcing is like having good students and poor students in a class, where good students let poor students copy their homework. From a basic business logic perspective, if such a good thing is used by Li Xiang Auto, it would enhance its competitiveness. Why let others copy? What role can these poor students play in the learning of good students during this copying process? Many poor students may choose to keep copying; what feedback effects does this have on Star Ring OS? Xie Yan: The automotive operating system is a very complex concept, encompassing vehicle control, intelligent driving, and the overall vehicle communication system. It connects various applications on top and various suppliers' hardware underneath. Relying solely on Li Auto to develop it would be very slow. If the industry wants to move to the next era and achieve the leap from Smart System to Intelligent System, realizing the future blueprint, it needs to open-source the system and allow more people to join; only then can we move better, more steadily, and faster. For example, when we tell a chip company that we need to add an IP (intellectual property design patent) to the ECU (controller) for better safety, if we don't open-source the system, the chip company will be unwilling to do it. Only by opening up the system will the chip company believe that this change can be used not only by Li Auto but also by other manufacturers, making them more willing to undertake this task. I personally believe that there is no absolute technological lead or absolute moat. Even if we remain closed-source, others will still develop their own automotive operating systems. For instance, iOS has always been closed-source, but Google quickly developed Android and made it open-source. In my previous work with many mobile operating systems, I found that technology is hard to maintain a long-term lead, especially at the underlying level, so we must keep moving forward. We hope to run faster with like-minded people; the faster we run, the more valuable it becomes. Question: So you want to attract more suppliers to join? Xie Yan: Suppliers and automakers. The more automakers that use the Li Auto Xinghuan OS, the more suppliers will actively embrace it. This story is already playing out in the mobile phone industry. For example, after Android was released, many phone manufacturers adopted it, and hardware manufacturers, including Qualcomm, actively embraced it. This is a resonance of frequency. The operating system is like a protocol, similar to how everyone conducts trade under the WTO agreement; the more people that join, the faster the (operating system) develops. Everyone can see the future direction of progress on this protocol, like looking at a common map moving forward. Question: If the Xinghuan OS ultimately aims to create a super AI agent, it means that the product's form will be controlled by a super AI agent, which will be managed by algorithms or software, including the vehicle system, intelligent driving, and battery. What is the underlying logic of this super AI agent? We have seen many safety incidents in recent years; how can we balance convenience and safety to avoid these at the most fundamental system level? Xie Yan: To clarify, we see that there are many agents in the car. Intelligence will become more dominant, but the Li Auto Xinghuan OS is not an agent; it serves these agents, which run on the Li Auto Xinghuan OS. Regarding the balance between convenience and safety, it is something that needs to be addressed at the application layer, which is the agent itself. For example, autonomous driving is an agent, similar to a driver; the safety issues must be resolved by the autonomous driving system, and the underlying system must also address these issues For example, end-to-end security and real-time performance. Even if the autonomous driving system has made a correct decision, if there is a long or uncertain delay between making the decision and executing a turn, brake, or some action, it can also cause safety issues. We hope to solve this through the operating system, aiming for more certain and shorter delays in task coordination, scheduling, and communication. Therefore, the operating system is not the agent itself. Q: Does this mean that the car itself is a super AI agent, or is it a collection of many agents inside the car? Xie Yan: This actually involves a bigger topic, which is "what will the future car look like." Li Xiang has also mentioned that we believe there will be many agents inside future cars. For example, you will have a driver, and there will be a butler responsible for entertainment, conversation, information retrieval, chassis adjustment, seat control, and temperature control. These could all be individual agents. Q: As a car company, it is relatively sensitive for other car companies to choose to use Li Auto's Xinghuan OS. So if the roles were reversed, what would Li Auto look for in choosing an open-source automotive operating system? Xie Yan: In the open-source model, we consider ourselves as a car company. If there is something particularly useful, under what circumstances would I be willing to use it? I think there are three conditions: First, it must solve the problems we need to address, and we must see the same vision or future. This is the primary driving force. Second, is there autonomy? If we use another automotive operating system that is delivered as a black box and we cannot see any code, then we cannot modify it. Even if we do not necessarily need to modify it, the code and solutions must be open and transparent. Third, we need to consider whether there will be constraints on commercial use. For example, some open-source projects can only be used for learning purposes, and if used commercially, a fee must be paid. I think this is not truly open-source. Therefore, we have considered these three aspects; we do not want to charge fees or restrict anyone's use. Q: Li Auto previously launched "Li OS" for intelligent driving, and "Xinghuan OS" also includes intelligent driving. So is this integrated together? Is there a conceptual distinction? Xie Yan: "Li OS" is a middleware for intelligent driving, specifically applied to Li Auto's intelligent driving application system, so we will not open this part. It belongs to the application layer, and the content of the application layer is subject to individual considerations and is related to their own commercial interests, decoupled from the system layer. Q: If a chip manufacturer custom-develops a chip based on "Xinghuan OS," what key parameters need to be met? Will "Xinghuan OS" have advantages over other general-purpose chips in the application process of custom chips? Xie Yan: First of all, there is not just one chip or one type of chip in the car. The intelligent driving system, vehicle control system, etc., require multiple chips, such as those from NVIDIA, ARM, TriCore, etc. The architectures are different. At the same time, there may also be multiple systems, such as multiple domain controllers that need to be linked together using Ethernet and other communication components So what we are targeting is not a single chip, but a group of chips. Secondly, for example, if a certain controller MCU (microcontroller) manufacturer adapts to the Aito OS, it means that it may save more resources compared to using software from other companies. The system's software and hardware as a whole can solve some issues related to security, performance, and efficiency, resulting in higher performance and better security. Finally, as an open-source system, chip manufacturers can start preparing for software porting during the chip development phase, which will significantly improve R&D efficiency and shorten the time for chips to be deployed in vehicles. Q: The scope of open source includes the chip part; what specifically does this refer to? Is it the underlying software or something else? Xie Yan: The platform of the operating system is the hardware platform, and the most core part of the hardware platform is the chip architecture. What is presented in the diagram refers to the chips that we can currently use, and of course, it also represents our welcome to more chip manufacturers to cooperate and join this open-source community to adapt to this system together. Reporter: To what extent does Aito OS go deep? Will it involve the underlying software of the chips? Xie Yan: The range of the underlying software of the chips is quite broad; for example, we will involve the kernel. However, for specific drivers and peripherals, they are still provided by the manufacturers. Q: Can you elaborate on how rapid adaptation is achieved? When communicating with some operating system manufacturers, they find that the deeper they go, the harder it is to adapt quickly because most operating systems are developed based on certain customized chips or specific mainstream products. Xie Yan: Rapid adaptation in this context refers to the ability to quickly deploy to vehicle products. For example, in a real situation, when Li Auto lacked a certain chip and needed to switch to another chip, the chip manufacturer proposed providing a usable system in six months, which is quite a long time. The reasons why Aito OS can achieve rapid adaptation are threefold: First, after becoming open-source, using the complete toolchain provided by Aito OS, car manufacturers and chip manufacturers can quickly collaborate without relying on others. Second, the "Aito OS" system itself has good decoupling. Previously, when we used other automotive operating systems, a lot of time during the adaptation process was spent on various adaptations between the system layer and the application layer. If there is a good decoupling layer between the application layer and the system layer, also known as the hardware abstraction layer, then during the system change process, there is no need to change the hardware or touch the application layer, which will reduce a lot of development and testing work. Third, Aito OS has a user-friendly toolchain. In the past, traditional toolchain information was not complete, and compared to other industries, the development of the toolchain in the automotive sector has been relatively slow; in the fields of mobile internet and server cloud, the tools are more modern, so we introduced a lot of technology to promote the speed of iteration. Q: The automotive industry may have been more accustomed to project-based or closed-source models for a long time. This time, Li Auto's open-source initiative is significant for the industry. In promoting the open-source of Aito OS and ecological construction, what do you foresee as the major challenges? What measures are in place to address them? Xie Yan: The challenge is how our system can be accepted and applied by everyone; for us, this is the only measure of success. The more people who can join in, the more it indicates that the system is correct and that the open-source approach is right. Achieving this goal is not easy. As mentioned by the media earlier, "no one has done this before." Besides the fact that it is indeed difficult to implement, there is also the issue of "timing." When a new thing emerges, there are always opportunities for change. For example, if there were no changes in the automotive industry, people would not change either; using the old things would be sufficient. The automotive industry would remain in the era of gasoline vehicles or simply in the era of basic intelligence and connectivity, only needing to meet the demand for users to connect and listen to music, and there would be no need for change. Q: The open-source community is relatively unfamiliar to the automotive industry, but what is the governance structure of the Li Auto Star Ring OS open-source project? Xie Yan: The governance structure is the same as that of a conventional open-source community: we envision a committee where active members join to collaboratively set versions and contribute together; there are some maintainers responsible for reviewing the code to ensure that each version is usable. Although the automotive industry is relatively unfamiliar with open-source, I believe many engineers in the automotive industry are also software engineers and hardware engineers, and they are not unfamiliar with open-source. Q: What is the follow-up rhythm for the Li Auto Star Ring OS to be implemented in vehicles? Can past models complete some technical upgrades through OTA? Xie Yan: The models we will release next will all be equipped with the Li Auto Star Ring OS, and some components have already been used in previous models. Because the Li Auto Star Ring OS involves not just a single piece of hardware, but as a system, it involves many controllers, some technologies can be used independently. Previous models will upgrade some usable technologies through OTA. Q: For users, where are the benefits? Xie Yan: The Star Ring OS is a foundational system and does not directly serve users; it may not be appropriate to compare it to Android, as Android and Windows are user operating systems. If we compare cars to robots, then the Star Ring OS is equivalent to a robot's internal foundational brain, cerebellum, and neural network system. It does not directly contribute services to you but supports the capabilities that underpin those services. To be more direct, everyone can feel that the update speed of Li Auto is very fast, from intelligent driving to the cabin, the system is becoming smarter, and the iterations will be even faster. If a system is good, hardware upgrades also become easier. For example, if a user buys a car and hopes to use it for ten years, but technology is developing so rapidly today, they may worry that in a few years, the car may not keep up. If the operating system is good enough, not only can the software be updated, but in the future, the hardware may also be upgradable. To achieve this goal, some innovations are needed at the foundational level, so we are supporting these iteration speeds, safety, and efficiency Question: Have you signed any limited agreements with other manufacturers, such as not making too many modifications to the current system? Xie Yan: We will not prevent users from making any modifications. Just like DeepSeek, there are no restrictions on modifications. There is a lot of room to move forward, and we need to be pragmatic in our product development. If something is not done well and it's our problem, we will change it; if it's not our problem, it will gradually be clarified. Question: Looking at a 3-5 year horizon, what stage does Li Auto hope the Xinghuan OS will reach? What are the goals? Xie Yan: Of course, we hope that Xinghuan OS will be accepted by more suppliers and car manufacturers, the more the better. Regarding goals, there are some technical objectives internally that will be gradually released. Risk Warning and Disclaimer The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this is at your own risk