Li Auto is making in-house battery development a central part of its vehicle strategy. Its executives said any battery manufacturer can participate, provided it meets Li Auto’s technical requirements and shares data and product specifications.
When the automaker first publicly addressed questions about its batteries in June, the debate centered on why it was developing them and whether they would be reliable. At a recent two-hour media briefing, the discussion shifted to how automakers’ involvement in battery development could reshape the supply chain.
Li Auto president Ma Donghui, senior vice president of vehicle electrification R&D Liu Liguo, and battery R&D head Liu Zhimin said at least three times that they welcomed any battery maker willing to work within the company’s system.
Li Auto began its broader rollout of batteries developed in-house with the new-generation Li L8 this year, followed by models including its flagship electric Li Mega and Li i9. The shift caused Li Auto to drop out of CATL’s top five customers in August, according to 36Kr.
Previously, CATL supplied the vast majority of Li Auto’s vehicle batteries, with Sunwoda serving as a second supplier. At its peak, Sunwoda’s share of Li Auto’s battery supply was only about 30%, according to 36Kr.
According to 36Kr, securing large shares of major customers’ orders has helped CATL maintain a market share of more than 40%. Li Auto’s shift therefore threatens CATL’s position with the automaker.
Under this model, Li Auto leads cell design and R&D, while battery manufacturers such as Sunwoda and CALB produce the cells under contract. At the pack level, Li Auto develops, designs, and produces the battery management system (BMS), thermal management system, and even insulation materials. CATL is absent from Li Auto’s current list of partners for batteries developed in-house.
Ma told 36Kr and other media outlets that in-house battery development was a firm strategic choice, with no alternative partnership model as a backup.
Liu Liguo added that every future supplier would have to develop customized products within Li Auto’s system. “Whatever the price, they must meet our requirements.”
That presents a structural challenge for CATL. Its Qilin battery for the premium market is an integrated pack solution: CATL integrates and supplies the cells, the complete pack, and components including thermal protection.
According to 36Kr, CATL has not previously supplied Qilin as separate components, partly to protect its IP, and it gives automakers limited access to battery production data.
That conflicts with Li Auto’s approach: leading cell development, making its own packs, and sharing data throughout battery production. To rejoin Li Auto’s supplier network on those terms, CATL would need to open up its battery system, share underlying data, and accept a cell-only supply arrangement. For a battery company whose market value once approached RMB 2 trillion (USD 298.2 billion), that would represent a substantial change in its business model.
Ma said Li Auto bears primary responsibility for vehicle safety. Without access to production data, it cannot fully understand the battery’s manufacturing and operating history.
The company sees three benefits: sharing data with cell makers can provide earlier safety warnings; making packs itself can improve coordination among cells, thermal management, electric drive, and charging; and common after-sales standards place responsibility with Li Auto, sparing customers disputes between the automaker and battery supplier.
CEO Li Xiang has said the strategy began in 2020. Li Auto’s battery R&D team now has more than 300 people. It has spent more than RMB 200 million (USD 29.8 million) annually on battery R&D over the past six years, bringing cumulative spending to more than RMB 1 billion (USD 149.1 million), excluding fixed assets such as laboratories.
Li Auto has also deepened its ties with Sunwoda’s electric vehicle battery business, Sunwoda EVB, through equity investment. It first invested in 2022. A further investment announced in September would make it the unit’s second largest shareholder upon completion.
Liu Liguo said batteries jointly developed with Sunwoda had been delivered in vehicles since 2022 and were installed in 420,000 vehicles at the time of the briefing.
“Their after-sales quality performance is every bit as good as the products of any battery supplier. For the products available today, I do not think there is any gap,” he said. That record has given Li Auto confidence to roll out its own batteries across its vehicles.
Liu Zhimin said the company’s R&D, control standards, and artificial intelligence-driven inspection systems had brought the rate of manufacturing defects escaping detection down to the parts-per-billion (PPB) level.
Other automakers are pursuing similar approaches. Xiaomi has its Longjia battery, while Huawei’s Harmony Intelligent Mobility Alliance (HIMA) has its “giant whale” battery. Leapmotor has moved into battery manufacturing through a joint venture.
Automakers want greater involvement in battery development and control over the full range of battery technologies. CATL, meanwhile, is building branded products that command a premium. The industry will have to accommodate both ambitions, even where they conflict.
The following transcript has been edited and consolidated for brevity and clarity.
Q: Why is Li Auto developing its own batteries? What are the main benefits?
Ma Donghui (MD): In 2020, Li Auto decided to develop its own battery electric platform. We believe new energy vehicle users care less about the battery brand than about the capabilities behind it, such as safety, range, and charging. In-house development is intended to address those issues systematically.
First, it connects data across components, the vehicle, and the user. We can understand battery usage and lifetime trends in advance. If production parameters become abnormal, we can issue an early warning. Ideally, we would identify problems before a battery is installed in a vehicle and delivered, instead of discovering risks only after it reaches the customer.
Second, the battery does not exist in isolation. It is closely connected to the thermal management, electric drive, and charging systems. Only by designing them as a whole can we truly optimize the experience of using the vehicle. We believe future intelligent EVs should not treat batteries, electric drive, and thermal management as separate modules. The allocation of functions and how they work together will change profoundly.
Third, it connects product specifications, service standards, and the after-sales service system. There are still difficulties in after-sales repairs. Automakers, users, and battery companies may apply different criteria to issues such as declining range. When a quality problem occurs, is it an integration issue, a cell issue, a pack issue, or a software issue? In-house development means that if quality standards are not met, users do not need to work out which part of the process is responsible. Li Auto takes full responsibility.
Q: What are Li Auto’s goals for in-house batteries? How large is the gap with CATL?
Liu Liguo (LL): Our battery investment is highly focused.
Other battery companies need to work on multiple systems, including energy storage and sodium batteries, while meeting different customers’ requirements for 2C, 4C, 5C, and higher charging rates. They have many product lines and invest heavily. We focus on just one category, 5C ultrafast charging. That focus is what allows us to develop the technology in depth.
The battery team does not work alone throughout R&D, either. Materials work focuses on cathodes and anodes. At the pack level, there are issues such as heat dissipation. The vehicle materials and thermal management departments also work together to support the effort.
Our investment in an individual product is no smaller than that of any other company.
As for actual products, batteries we developed and specified in-house, with Sunwoda manufacturing them under contract, have been on the market since the first-generation Li L8 and Li L7. They are now installed in 420,000 vehicles, and their after-sales quality performance is every bit as good as that of any battery supplier’s products. For the products available today, I do not think there is any gap.
But in emerging technologies, such as the ability to develop solid-state batteries, there is indeed still a fairly large gap between us and CATL. We acknowledge that openly.
CATL aims to lead the entire field. Our current strategy remains driven by the product experience. Our strategies and scale of investment differ, so naturally the technology we have developed differs too. But for today’s products, our past R&D and validation through deliveries of 420,000 vehicles mean our performance is every bit as good.
Q: Many automakers are putting their own batteries into vehicles, but most have yet to answer one question: is in-house development a backup or the only future route? If this model of in-house development and contract manufacturing expands, will it include all battery makers, including CATL? Or could CATL remain a separate, parallel route if it lowers its prices?
MD: I am not entirely clear on other automakers’ thinking, but Li Auto’s direction is very firm. Starting with the all-new Li L8, our vehicles will use batteries developed by Li Auto. That strategic direction is definite. It is not a backup, and we are confident that the overall performance of our batteries will be better than before.
We now have the technical capabilities and our own battery pack production lines. The company has also reached a point where its investment capacity can support this strategy. For us, this is a clear strategic choice. There is no question of it being a backup.
LL: The pack is closely integrated with the vehicle. In-house development allows us to standardize the pack design effectively across the vehicle platform. With a common pack, we can specify customized requirements for cells.
Whatever the supplier, it will need to develop customized products according to our system and dimensional standards. Whatever the price, it must meet our requirements. If a supplier can meet our three requirements for integration, we are of course very open to working with any strong partner in the industry.
Q: Li Auto is still smaller than CATL and BYD. Without the same scale, how can in-house batteries offer an advantage in cost and quality?
MD: On scale and cost, I would first explain how we think about cost. Battery costs have five main components: production line depreciation, manufacturing expenses, raw materials, logistics and warehousing, and R&D investment. We are not developing batteries to squeeze the price of one part of the process, still less to substitute cheaper materials. That would not fit Li Auto’s values. We see opportunities to reduce costs by improving efficiency across the entire process while preserving performance and quality.
First, production line investment and depreciation: suppose a production line costs RMB 1 billion, depreciated over several years. By standardizing cell platforms, we can reuse production lines as much as possible. Higher utilization reduces the depreciation cost allocated to each unit.
Second, manufacturing expenses: Li Auto has an intelligent manufacturing team and the Lianshan system to monitor production and analyze data. That supports quality warnings while identifying waste and improving yield.
Third, raw materials: in the past, we largely had to accept fluctuations in raw material prices. With in-house development combined with either our own manufacturing or contract manufacturing, we can take a more active role in upstream materials and the associated risk management, improving our ability to manage cost fluctuations and maintain a stable supply.
Fourth, logistics, warehousing, and transport: a common battery system makes storage, packaging, tooling, and transportation more efficient, reducing waste from duplicate storage and transport.
Fifth, R&D investment: R&D ultimately contributes to the cost. Previously, suppliers paid for R&D and reflected that in the purchase price. Now Li Auto makes the R&D investment, while partners focus on manufacturing and processes. That creates a clearer division of work across the chain.
The goal is to reduce waste and improve efficiency at every stage without compromising performance, quality, or safety.
Q: How does Li Auto plan to standardize cell platforms?
Liu Zhimin (LZ): At the cell level, we currently have two solutions: 5C ternary lithium and 5C lithium iron phosphate. Each category has no more than two cell types.
Q: Reports have put Li Auto’s battery R&D spending over the past five years at about RMB 500 million (USD 74.6 million). What is the actual cumulative investment?
LZ: The RMB 500 million cited in reports largely refers to only part of the R&D expenditure, such as laboratory materials and testing and validation. Li Auto spends more than RMB 200 million a year on battery R&D and has spent more than RMB 1 billion over the past six years. There are also fixed assets and other related investments. The overall scale of R&D investment is substantial.
Q: Some figures suggest that, as of early 2026, Sunwoda’s electrochemistry R&D team had about 10,000 people, CATL had more than 30,000 relevant engineers, and Li Auto’s battery R&D team had about 300. How do you view this difference between automakers and battery companies?
LL: Headcount is closely related to the size of a company’s business and the scope of its R&D. A battery company may serve several automakers while also covering energy storage, different material systems, and platform technologies. Naturally, it needs more people.
For an automaker, what matters more is not the absolute number of R&D staff, but the depth of the work, its focus, and whether it has clear goals for users and products.
Q: BMW also develops batteries in-house and contracts out manufacturing, but consumers are more accepting of it. How can Chinese companies address the perception challenge of removing a supplier brand when developing key components themselves?
LL: There are several reasons, including brand history and established market perceptions. Newer companies have existed for less time, and that does have an effect.
But let me give a specific example. Li Auto began defining charging rate requirements for its battery electric platform in 2018 and 2019. When we spoke to battery makers, the industry generally thought 2C was enough. Leading suppliers told us 4C was impossible, but we had already set a target above 4C.
In that context, Li Auto led the development of early demo products and validated the feasibility of the entire technical approach. That allowed us to continue advancing toward 5C. We have invested substantial R&D resources in both batteries and vehicles, from the start of in-house development in 2020 to the announcement of our “dual-energy strategy” in 2023 and the full rollout of 5C batteries and the ultrafast charging network in 2024.
In explaining our brand, we have certainly faced plenty of skepticism. It takes time for users to develop an understanding of batteries developed by newer automakers.
MD: I am not worried at all. In 2017 and 2018, people might not have heard of Li Auto and would probably not have considered it when shopping for vehicles above RMB 200,000 (USD 29,800) or RMB 300,000 (USD 44,700). But a decade has passed, and as Li Auto and other new Chinese automakers have continued to improve their products, our share of the domestic market above RMB 200,000 has already exceeded that of BBA (BMW, Mercedes-Benz, Audi) in some regions. I always emphasize that brands are not fixed forever. Value for users is what stands the test of time.
Q: What is Li Auto’s rate of cell manufacturing defects escaping detection, measured in PPB?
LZ: Before discussing the manufacturing PPB, or parts-per-billion, measure, let me briefly explain how we understand manufacturing consistency.
A complete product plan usually has three stages. First, we develop a complete technical solution around the product requirements. That solution connects all the product’s technical characteristics. Second, we translate those technical characteristics into engineering characteristics in the engineering plan.
Third, we ensure that the engineering characteristics correspond precisely to specific control plans in manufacturing. At the cell level, we have strict control standards and control items: more than 8,000 standards for cells and more than 1,000 at the pack level, for a combined total of more than 9,400. These ensure consistency in battery manufacturing.
Beyond the control plan system, we have deployed AI-driven inline CT inspection throughout the production process, along with AI-driven inspection using CCD (charge-coupled device) cameras with liquid lenses and other systems. These can effectively monitor abnormalities during manufacturing.
We also connect algorithms to Li Auto’s Lianshan system, using AI to identify abnormal trends and continuously improve and iterate. At present, our consistency meets product requirements, and the cell manufacturing defect escape rate has reached the PPB level.
Q: What specific technical coordination do Li Auto’s in-house batteries enable at the vehicle’s electronic control level?
LL: Technical coordination among the cells, the pack, the BMS (battery management system), and the vehicle is already extensive. Let me give a few typical examples.
The first is discharge power at low temperatures. Under the traditional model of working with battery companies, the supplier would provide a power map setting discharge power limits at different temperatures, and the vehicle could only be calibrated according to that map. This creates problems, especially for lithium iron phosphate batteries, which can deliver insufficient power or even leave the vehicle unable to move at low temperatures.
Through in-house development, we have our own APC (adaptive power control) and ATR (adaptive trajectory reconstruction) algorithms, allowing us to calculate SOC (state of charge) and the power map more precisely. There is a delay of tens of milliseconds between a driver pressing the accelerator and the demand signal being issued. We use that window to provide a feedforward control signal to the battery. Close coordination between the vehicle control algorithm and the battery system can significantly reduce power shortfalls at low temperatures. That is a direct practical benefit of in-house development.
Second, beyond power, we have also achieved more integrated control in areas such as thermal management. In-house development gives us basic production-stage data, including the capacity of every cell in the battery system. That enables more precise, coordinated management, particularly as battery health changes after long-term use.
KrASIA features translated and adapted content that was originally published by 36Kr. This article was written by Xu Caiyu for 36Kr.
Note: RMB figures are converted to USD at rates of RMB 6.71 = USD 1 based on estimates as of October 13, 2026, unless otherwise stated. USD conversions are approximate and, where appropriate, rounded for ease of reference. They may not fully match prevailing exchange rates.