In the first half of 2026, the penetration rate of NEVs (New Energy Vehicles) further increased, surging to approximately 64.5% in July. The penetration rate of Level 2 combined driving assistance features in new domestic passenger cars has reached 70.5%, with the penetration rate of NOA (Navigate on Autopilot) features reaching 34.2%. The cost of intelligent configurations per vehicle for domestic brands has dropped by over 60% compared to three years ago... As "equal pricing for ICE and NEVs" becomes the norm and intelligence shifts from a novelty to a necessity, a key signal is clear: behind these figures lies a shifting growth logic.
Over the past decade, the growth of China's auto industry has relied on "volume expansion"—capacity ramp-up, channel sinking, and trading price for market share. It can be said that electrification was the biggest technological dividend in this stage, solving the "zero to one" problem and allowing China's auto industry to sit at the global table for the first time.
However, the competitive landscape in 2026 has changed. As the NEV penetration rate crosses the 50% threshold, and L2+ intelligent driving shifts from an optional feature in a few models to a standard factory installation in most new cars, a simple "have or not" can no longer constitute a competitive advantage. The driving force for growth is shifting from "the popularization of electrification" to "the deepening of intelligence," from "cost leadership" to "technological premium," and from "single-point breakthroughs" to "systematic operations."
Today, the very definition of the technological dividend is being rewritten. It is no longer about the domestic substitution of a specific component, nor about leading by a few seconds in a certain parameter, but rather—who can use technology to redefine user experience, who can use technology to reconstruct the cost structure, and who can use technology to establish pricing power in the global market. These three questions determine where the growth ceiling for China's auto industry lies in the next decade, and also what the answer to "going far" is.
Self-Reliance in Core Technologies: From "Choke Points" to "Counter-Positioning"
To answer the above three questions, we must first return to the most fundamental physical foundation of car manufacturing—the three-electric system (battery, motor, and electronic control).
In 2026, the monthly sales share of NEVs in China broke through 60% for the first time, with installed capacity of power batteries continuing to surge. But the significance of the numbers goes beyond scale itself—the independent right of Chinese enterprises to choose their technological routes is shifting from "following" to "defining," which is the real reason why the three-electric system can truly be called the "cornerstone."
In the battery sector, the duopoly of CATL and BYD is a microcosm of Chinese enterprises reconstructing global industrial discourse.
According to data from Gasgoo Research, in the first half of 2026, among the installed capacity of power batteries for domestic NEV passenger cars, CATL ranked first with 100.7GWh and a 42.7% market share, followed closely by BYD FinDreams Battery with 53.8GWh and a 22.8% share. The combined market share of these two leading enterprises exceeded 65%. CALB, and Gotion High-Tech ranked third and fourth, with a share gap of less than 0.6 percentage points.
In addition, from January to June 2026, the global installed capacity of power batteries reached 608.5GWh, a year-on-year increase of 20%, among which seven Chinese enterprises collectively held a 72.4% market share.
However, true technological self-reliance extends beyond batteries. In the electric drive system sector, the application of 800V high-voltage platforms and SiC (Silicon Carbide) power devices is accelerating. In 2026, the penetration rate of 800V high-voltage platform models in China's NEVs has reached around 20%. The domestic substitution of SiC devices has also achieved breakthroughs, with automotive-grade products from enterprises such as Xinlian Integration and StarPower entering the supply chains of mainstream automakers, gradually breaking the monopoly of international giants.
If the three-electric system is the "basic skill" of electrification, then intelligent driving is the "commanding height" of intelligence. In this field, Chinese enterprises have forged a "Chinese path" distinct from Tesla's "pure vision" route—multi-sensor fusion, especially the deep integration of LiDAR (Light Detection and Ranging) and visual algorithms.
The 2026 P3 China Intelligent Driving Evaluation Report shows that Huawei Qiankun ADS V4.1 ranked first with a score of 4.46, followed closely by XPeng XNGP VLA 2.0 with 4.33, and Li Auto AD Max ranked third with 4.06. Although all three are in the first tier, their technological paths have different focuses.
The divergence of these technological routes is rooted in the particularity of China's road environment—complex traffic participants, road complexity, and changeable weather, which all force Chinese enterprises to invest more R&D resources in algorithm robustness and system redundancy.
According to relevant authoritative data, in the first half of 2026, against the backdrop of a nearly 20% year-on-year decline in passenger car deliveries in the Chinese market, the sales of factory-standard NOA models grew by 51.23% year-on-year, and the penetration rate of intelligent driving domain control chips rose to 37.21%—the colder the market, the more intelligence becomes a rigid demand.
Intelligent driving chips are the "hardest bone to crack" in technological self-reliance. And Horizon Robotics' breakthrough path provides a highly referential sample for this dilemma.
As of May 2026, the cumulative shipments of Horizon Robotics' Journey series chips have exceeded 11 million units, with about 1.95 million units shipped from January to May 2026. More importantly, in April 2026, Horizon Robotics released China's first cabin-driving fusion vehicle intelligent agent chip "Starry", adopting the 5nm automotive-grade process of TSMC (Taiwan Semiconductor Manufacturing Company), with a computing power of 650 TOPS. It integrates the four domains of intelligent driving, cockpit, instrument, and vehicle control into a single chip, reducing space occupation by 50% and lowering the cost per vehicle by 1,500 to 4,000 RMB.
The significance of this breakthrough goes far beyond the accumulation of commercial orders—it means that China's auto industry has gained the confidence to confront foreign giants head-on on core computing platforms.
Yu Kai, founder and CEO of Horizon Robotics, frankly stated at the 2026 China Automotive Chongqing Forum: "We are following the strategy of encircling the cities from the countryside, and ultimately we must seize the heartland of the market." "Squeezing" orders out of NVIDIA and Mobileye requires not only meeting technical standards but also building long-term trust in ecosystem construction and customer stickiness.
"Spillover" of Technological Dividends: From Complete Vehicles to the Industrial Chain
When technological self-reliance accumulates to a certain extent, its dividends begin to penetrate deep into the industrial chain, forming a "multiplier effect" that transcends individual enterprises.
BYD's "vertical integration" model is the best sample for understanding the spillover of technological dividends. In the first quarter of 2026, BYD FinDreams Battery topped China's battery PACK installed capacity with 451,000 units and a 25.4% market share. Behind this advantage is BYD's integration of the entire industrial chain closed loop from lithium mining to complete vehicle production, making its battery costs 15%-20% lower than the industry average, providing the foundation and support for the large-scale spillover of technological dividends.
However, BYD's full-stack in-house R&D is not the full picture of the industry. As technology matures and capacity expansion continues, its spillover effect is increasingly evident. FinDreams Battery is externally supplied to automakers such as Changan and FAW, with rapidly expanding scale, and the independent value of the battery business segment is accelerating its release.
Image source: BYD Auto
At the same time, BYD's "make-up classes" in the field of intelligence are also accelerating—in 2025, its self-developed "Eye of God" high-level intelligent driving system achieved mass production and installation in vehicles. Entering 2026, BYD further accelerated the in-house R&D of intelligent driving chips, releasing its first 4nm process chip "Xuanji A3", aiming to close the final gap from the three-electric systems to intelligent driving algorithms.
Different from BYD's complex layout of "full-stack in-house R&D and vertical integration," new forces such as NIO, XPeng, and Li Auto have shifted their in-house R&D focus to the fields of intelligent chips and algorithms, choosing a compromise path of "moderate external procurement + core in-house R&D." In 2026, NIO's 5nm Shenji NX9031 chip took the lead in achieving large-scale delivery, with cumulative shipments exceeding 250,000 units; XPeng's second-generation VLA (Vision-Language-Action) large model achieved end-to-end direct generation from "visual signals to action commands," and reached a cooperation with Volkswagen, which is expected to contribute over 5 billion RMB in profit when supporting Volkswagen models in 2026.
The most intuitive manifestation of the industrial chain spillover effect is the rise of China's Tier 1 suppliers. According to data from Gasgoo Research, since 2026, in the core components market for electrification of NEVs in China, the FinDreams series has ranked first in multiple segments such as battery PACK, BMS (Battery Management System), drive motors, and electronic control. Local suppliers such as Huawei Digital Power, Huichuan United Power, and CRRC Times Semiconductor have performed outstandingly in their respective tracks. The shares of international manufacturers such as LG Energy Solution, Infineon, and Denso have been significantly compressed, accelerating domestic substitution.
In the field of intelligent chassis, Bethel Automotive Safety Systems has achieved a mass production breakthrough in the One-Box brake-by-wire system, breaking the monopoly of international giants such as Bosch and Continental; the air suspension systems from Chinese local suppliers like Baolong Automotive and Konghui Technology have reduced the cost to a range affordable for vehicles priced under 300,000 RMB. According to authoritative data, in 2026, the installation rate of air suspension for passenger cars in the Chinese market has exceeded 10%, an increase of nearly 15 times compared to 2020.
The spillover of technological dividends has not stopped at the supply chain itself; it has also penetrated into broader industrial scenarios. CATL is transforming from a pure battery supplier to a full life-cycle service provider of "energy storage + battery swapping + recycling," and its "Chocolate Battery Swapping" model has been implemented and operated in cities such as Xiamen and Hefei; BYD's integrated optical storage and charging solution, NIO's battery swap station network, etc., are all extending automotive technological dividends to the energy infrastructure field.
The profound significance of these explorations lies in that they expand the boundary of technological dividends from "transportation tools" to "energy nodes," extending from one-time hardware sales to continuous service operations. Technology is more than just building cars—this is the true essence of industrial chain spillover.
However, technological self-reliance and a complete industrial chain do not mean resting on one's laurels. Zhang Hong, Deputy Secretary-General of the China Automobile Dealers Association, pointed out that the true watershed for the industry lies in the ability to transform technology into sustainable business models. "Core technology determines the lower limit, and systematic capability determines the upper limit. Single-point technological breakthroughs are difficult to support long-term development. Only by deeply synergizing chips and algorithms with complete vehicle architecture, supply chain, and manufacturing systems can they be truly transformed into user experience and brand moats."
From Technological Self-Reliance to Global Definition
If self-reliance is "practicing internal kung fu" and industrial chain spillover is "strengthening the ecosystem," then technology going global is the critical leap for China's auto industry from "follower" to "definer."
In 2026, this transition is moving from individual cases to the norm: from product exports to localized capacity, and then to technology licensing and standard output, the narrative of China's auto industry going global is being rewritten. The core driving force of this qualitative change is precisely technology.
Despite the continuous escalation of anti-subsidy tariffs imposed by the EU on Chinese electric vehicles, the market share of Chinese brands in the pure electric vehicle market across 18 Western European countries climbed to 14.2% in the first five months of 2026, hitting a record high. This marks a record high. Tariff barriers have not eroded their competitiveness, but instead forced Chinese automakers to accelerate global capacity layout. BYD's Hungary factory is about to commence production, while Chery, Geely, and SAIC Motor have simultaneously set up operations in Spain, further compressing planning and production launch times. At the same time, factories of Chinese automakers in markets such as Thailand and Brazil have been successively put into operation, and China's auto industry is moving from "exporting globally" to "producing globally."
Deeper than capacity going global is the exploration of technology licensing models, such as the cooperation between Leapmotor and Stellantis Group, which can be called a model.
In 2023, Stellantis invested 1.5 billion euros to acquire about a 20% stake in Leapmotor, establishing a joint venture "Leapmotor International," with Stellantis responsible for the export and sales of the Leapmotor brand in other global markets. The core of this model is that Leapmotor exports its core capabilities, such as its self-developed central integrated E/E (Electrical/Electronic) architecture and CTC (Cell to Chassis) battery-chassis integration technology, to Stellantis, while Stellantis provides its globalized manufacturing, sales, and service network.
The advantage of this model lies in risk-sharing and benefit-sharing, avoiding the high costs and uncertainties of a single enterprise exploring the global market alone. However, its challenges are equally obvious—where is the boundary of technology licensing? How to protect intellectual property? How to balance the discourse power between the two cooperating parties? The answers to these questions determine how far technology going global can go.
"This form of 'reverse technology export' is not a trend-chasing move," analyzed Zhang Hong. "It is highly likely to be a long-term trend, but for now, it will not be smooth in all areas of the industry."
In his view, there are multiple driving forces behind this trend: domestic market saturation, declining capacity utilization rate, and price wars squeezing profits, forcing enterprises to seek incremental growth overseas; at the same time, trade barriers are also forcing enterprises to shift from pure exports to exporting technology, architecture, and standards; more importantly, Chinese technology has genuine competitiveness, "not just cheap, but with strong engineering, fast iteration, and complete supporting facilities." In addition, the demand for a "second choice" in overseas markets is also rising. "Many countries do not want to be bound by a single European and American technological system. The Chinese solution wins with its high cost-performance ratio, flexible customization, and fast delivery."
A higher level of technology going global is the competition for standard discourse power. China's ChaoJi charging standard is gaining increasing international recognition. With its strong compatibility and wide power coverage, it is expected to become an important option for the next generation of global charging standards.
However, the path of standard output is not smooth. The EU is trying to build barriers with its own technical standard system—from charging interface standards to V2X (Vehicle to Everything) communication protocols, from battery specifications to carbon footprint accounting methods, the compatibility game between Chinese solutions and existing European standards is becoming increasingly fierce. The battle for standards is, in essence, a battle for future industrial dominance.
Zhang Hong also expressed regarding this, "It will not be a smooth journey. Compliance thresholds are getting higher and higher. Data, privacy, export controls, carbon footprint—any of these checkpoints could encounter blockades. In addition, localization is the real test. Having technology alone is not enough; you also need to understand local talent, channels, and user habits."
As Zhang Hong said, the sustainability of technology going global ultimately depends on the depth of localization capabilities, but localization is not easy. Issues such as cultural differences, management conflicts, labor union systems, and talent mobility all require enterprises to gradually adapt in practice.
"A large export volume does not equal strong globalization capabilities," Zhang Hong further emphasized. "Overseas after-sales, spare parts, finance, residual value of used cars, data compliance, and local R&D are all still under construction. Selling cars is just the beginning; what we rely on in the next decade is taking root in the system." He also pointed out, "For the same quality flaws, traditional luxury brands call them 'isolated cases,' while domestic brands are easily amplified into 'common problems'—this is the gap in brand fault tolerance."
Conclusion
The technological dividend is not a "get out of jail free card" that works once and for all, but a "dynamic asset" that requires continuous investment. From the self-reliance of the three-electric system to the breakthrough in intelligent driving chips, from the spillover of the industrial chain to the game of technology going global, China's auto industry is undergoing a transformation from "quantity" to "quality."
This is not a quick decision battle, but a protracted war that requires patience and perseverance, determining how far we can go and how steady we can stand.