On the afternoon of September 15, a joint announcement marked a pivotal shift in the partnership between Seres and Huawei, ushering in a new era that would test Seres's independence.

The announcement detailed that Seres would immediately assume leadership in product definition, design, brand marketing, channel retail, and service systems for AITO, with Huawei's consumer business group providing supportive roles. Concurrently, AITO would transition further towards an 'exclusive franchising' model. In addition to AITO's company-owned and franchised stores, select outlets under the Harmony Intelligent Mobility Alliance (HIMA) would also be transferred to Seres.

This strategic realignment does not extend to other brands such as Luxeed, Stelato, Maextro, and Shangjie, which will continue to operate under Huawei's comprehensive leadership model.

Despite the announcement downplaying any notions of a 'split' and affirming AITO's place within the HIMA ecosystem, market reactions were telling. On September 15, Seres' A-shares closed at 45.46 yuan, down 5.09%, while its Hong Kong shares fell to HK$34.68, a 6.12% decline, marking a new low since its listing on the Hong Kong Stock Exchange in November 2025.

As Huawei steps back into a supportive role, the more pressing question is whether Seres can leverage its past accumulations to forge its own competitive edge, rather than focusing on the speculation of a 'split' between Seres and Huawei.

I. An Inevitable Evolution

The evolution in the cooperation model between Seres and Huawei did not occur overnight. Over the past two years, both sides have been gradually redefining their control over AITO.

In July 2024, Seres announced its intention to acquire 919 AITO-series trademarks and 44 design patents from Huawei and its affiliates for 2.5 billion yuan, signaling the beginning of AITO's brand assets reverting to Seres.

A month later, Seres revealed plans to invest 11.5 billion yuan to acquire a 10% stake and a board seat in AITO's parent company, with payment due by September 2025.

Sales and service operations for AITO have long been a joint effort. According to LatePost, Huawei Experience Centers were responsible for product display, test drives, and customer acquisition, while Seres-managed user centers handled sales, delivery, and after-sales services. In 2023, both parties established a joint sales and service working group.

Starting in 2024, Seres began actively recruiting luxury brand dealers to sell AITO vehicles. At that time, brands like Luxeed and Stelato primarily relied on Huawei and China Post's sales networks, with partner automakers unable to establish their own. Seres's proactive approach signaled its shift towards independently acquiring customer traffic and building its own network.

By the summer of 2025, sales control for Luxeed, Stelato, and Shangjie had gradually been transferred to their respective partner automakers, although product definition and brand marketing remained under Huawei's purview. AITO's recent handover represents a more extensive shift, with Huawei retreating to a 'technology-enabling' position.

Why is AITO undergoing this transition?

Several factors that previously supported deep integration between the two sides are now evolving.

A technical launch event in March of this year highlighted these changes.

On March 4, Huawei unveiled its new-generation dual-optical-path image-level 896-line LiDAR, which was simultaneously launched on the Maextro S800 and AITO M9. However, before AITO M9 owners could celebrate, other HIMA partner brands quickly announced the adoption of related technologies.

Initially, AITO enjoyed the implicit privilege of being the 'first landing point for Huawei's latest technologies.' However, HIMA has now transitioned from the 'AITO era' to a phase where five brands operate in parallel. Huawei's collaborations with automakers such as Chery, BAIC, JAC, and SAIC continue to expand, with its technology, channel, and ecosystem capabilities no longer serving just one brand.

More direct conflicts arose among AITO M7 owners.

On March 23, it was announced that the 896-line LiDAR version would be extended to the M7. At that time, the 2026 M7 had only been on the market for a few months, with some owners having just taken delivery 1–2 months prior. This announcement triggered strong dissatisfaction among some owners because the M9 had previously offered a paid upgrade to the 896-line LiDAR, while their newly purchased M7 did not support it. Subsequently, AITO responded that it was actively seeking solutions.

By the time the AITO M6 was launched in April, the 896-line dual-optical-path LiDAR had become standard across all models.

The issue thus evolved from a routine product iteration into a question of user rights and brand value.

For automakers, rapid iteration itself is not the problem. The challenge lies in determining who sets the product rhythm, manages user expectations, and ultimately bears the cost of maintaining brand relationships when different models exhibit significant hardware generational gaps within a short timeframe.

This underscores a structural issue in the old cooperation model: Huawei controlled a large portion of product and intelligence capabilities, while Seres faced operational pressures from sales, delivery, and user service for specific models.

As product updates accelerated and the number of partner brands grew, the demands on product rhythm, resource allocation, and user operations became increasingly complex for both sides. The old model was no longer sustainable.

II. Why Now for the Transition?

On one hand, AITO is no longer the nascent brand that relied on Huawei's support to gain a foothold.

When AITO first launched in 2021, Seres's role was more akin to a manufacturing and delivery contractor: Huawei defined the products, while Seres built the vehicles. By 2026, however, AITO's cumulative deliveries had surpassed one million units, with the M9 firmly established in the market segment above 500,000 yuan.

On the other hand, Seres's own operational situation had reached a point where adjustment was necessary.

The automotive industry is becoming increasingly competitive. Data from the China Passenger Car Association shows that retail sales of passenger cars in the first half of 2026 exceeded 8.7 million units, down 20.2% year-on-year.

In the first half of the year, Seres managed to maintain its sales volume. New energy vehicle sales reached 178,800 units, up 3.87% year-on-year, including 160,800 Seres-branded vehicles, up 5.6% year-on-year.

However, due to a decline in the sales proportion of high-value models like the M9 and M8, average revenue per vehicle fell. As a result, Seres's operating revenue in the first half of 2026 decreased by 7.87% year-on-year to 57.493 billion yuan.

In the following two months, Seres's vehicle sales were 20,480 and 20,652 units, respectively, with declines of around 50%, while overall new energy vehicle sales fell by more than 40%.

The intense competition in the auto market is also reflected in technological innovation. As industrial transformation deepens, automotive technological innovation is entering a new phase of parallel breakthroughs across multiple paths, with L3 automation being a necessary step for large-scale commercialization of autonomous driving in passenger vehicles. Against this backdrop, some inventory materials, production equipment, and even research and development outcomes may no longer be applicable to new models.

As a result, Seres recognized 1.82 billion yuan in asset impairment losses in the first half of 2026.

Coupled with rising costs, Seres's profit margins were further squeezed.

Zhang Xinghai, Chairman of Seres, previously stated publicly that the price of memory chips had risen from about 20 yuan to nearly 100 yuan, while the price of lithium carbonate had increased from 80,000 yuan per ton to 180,000 yuan per ton. As a result, the manufacturing cost per AITO vehicle rose by approximately 15,000 to 20,000 yuan.

In the first half of the year, Seres's comprehensive gross margin was 23.3%, down 5.63 percentage points year-on-year; the gross margin for complete vehicles was 22.16%, down 6.36 percentage points year-on-year. Net profit attributable to the parent company turned from a year-earlier profit of 2.941 billion yuan to a loss of 1.717 billion yuan, including a quarterly loss of 2.471 billion yuan in the second quarter.

Under these circumstances, Seres had to change. The adjustment in its cooperation model with Huawei also aimed to alter the cost structure in its profit and loss statement.

According to Seres's Hong Kong stock prospectus, from 2022 to 2025, the amounts paid by Seres to its largest supplier were 5.8 billion yuan, 7.2 billion yuan, 42 billion yuan, and 56 billion yuan, respectively, accounting for 14.5%, 17.4%, 30.2%, and 33.78% of total procurement.

This largest supplier is widely believed to be Huawei.

Over four years, a cumulative total of 111 billion yuan was paid for intelligent cockpits, advanced driver-assistance systems, and other hardware and software, as well as Huawei's nationwide store channel services and brand marketing support. What Seres purchased was not just hardware but also the 'ability to sell cars.'

Contrary to what many believe as 'profit-sharing,' Huawei's role is that of a supplier and channel service provider.

In the past, Seres relied on Huawei's brand, channels, and intelligence capabilities to quickly gain sales volume. During the high-growth phase, these investments could be absorbed by the expanding scale.

Now, with fiercer market competition, changes in product mix, and rising costs and R&D investments, continuing the original model would require Seres to bear increasing operational pressures.

After regaining control over channels and operations, Seres must also shoulder new costs for channel construction, user operations, and product management. Whether the saved expenses can cover the new investments remains to be seen.

III. Can Seres Take the Helm?

So, can Seres make its vehicles more profitable after taking the helm?

First, let's consider what users are buying when they purchase an AITO vehicle.

According to a survey report on the first batch of AITO M9 owners released by JL&F Consulting, the reasons prompting users to purchase the AITO M9, listed in order of mention frequency, are as follows: intelligent driver-assistance functions rank first at 52.4%, followed by brand and intelligent cockpit, both at 43.5%, safety at 41.9%, comfort at 37.9%, and configuration at 33.9%.

Let's first look at intelligence.

Seres's strategy for intelligence is 'integration' rather than 'in-house research and development.' The company emphasizes that its Magic Cube platform has achieved 'simultaneous simulation, simultaneous calibration, and simultaneous verification' for Huawei's intelligent driving, HarmonyOS cockpit, and the vehicle's overall architecture. Its core capability lies in integrating different technological solutions into the complete vehicle.

After the transfer of control, Seres's demand for Huawei's intelligent technologies will not disappear; in fact, it may become more rigid as Seres independently assumes responsibility for product definition.

The situation with brand perception is more complex.

It cannot be denied that some owners purchase AITO vehicles because of Huawei, even believing that Seres is merely responsible for manufacturing the cars on Huawei's behalf.

To reverse this perception and convince users that 'Seres's AITO' is still worth buying, Seres will need time and products to prove itself.

Seres's strengths lie in powertrain and complete vehicle engineering.

In recent years, Seres has continuously increased its R&D investments. In 2025, Seres's R&D expenditure reached 12.51 billion yuan, up 77.4% year-on-year, accounting for 7.6% of revenue. In the first half of 2026, R&D spending was 7.007 billion yuan, up 34.8% year-on-year, accounting for 12.19% of revenue.

These funds were primarily invested in mass production technology areas such as the Magic Cube technology platform 2.0, a new-generation super extended-range system, and complete vehicle intelligent safety architecture, as well as in cutting-edge business layouts like embodied artificial intelligence.

In terms of powertrains, Seres has completed development of its fifth-generation 2.0T super extended-range technology, achieving 'vehicle-battery-motor-charging' integration. Leveraging 800V and above high-voltage architectures, octa-integrated electric drives, and a nationwide supercharging network, Seres continues to optimize energy efficiency per kilowatt-hour, significantly enhancing user travel efficiency. In 2025, Seres's extended-range system business captured a 37.5% market share in China, ranking first.

In complete vehicle engineering, Seres has built its capabilities around the Magic Cube technology platform 2.0, completing a 'central vehicle control + zonal control' electronic and electrical architecture and an eleven-fold safety redundancy layout. Relying on a smart factory with 100% automation of key processes and a 10,000-ton integrated die-casting process, Seres has formed a synergistic complete vehicle engineering capability encompassing platform development, safety, and manufacturing.

The AIVA brand serves as a litmus test for validating this capability distribution.

In 2026, Seres divested its long-unprofitable proprietary brand 'Seres Blue' and introduced external investors to complete a capital increase of approximately 6.671 billion yuan, renaming it 'Saido Technology' and launching the new brand 'AIVA.'

AIVA's technological route is entirely different from AITO's: it adopts the AutoX solution for intelligent driving, integrates the Volcano Engine and Doubao large model for the intelligent cockpit, relies on CATL for the energy system, while Seres contributes complete vehicle manufacturing, supply chain, quality systems, and engineering experience. The first mass-produced model, the AIVA ME7, is scheduled to debut in 2026, targeting the mainstream market segment above 200,000 yuan.

The significance of AIVA lies in it being the first time Seres has supported a new brand using its own engineering capabilities and manufacturing systems without Huawei's technological backing.

Should AIVA achieve success, it will serve as a compelling demonstration of Seres' comprehensive vehicle engineering prowess, affirming its ability to independently support a brand. Conversely, in the event of failure, Seres will face even more daunting challenges, including its technological reliance on Huawei for intelligence solutions and the persistent "Huawei shadow" that looms over its brand perception.