A New Option Emerges for Seres?

At the start of 2026, Seres’ performance on the A-share market presents a puzzling divergence.

Its financial statements remain polished, delivery figures are undeniably strong, yet the stock price tells a different story. Even after the AITO brand topped monthly sales rankings in January 2026, posted more than 10% growth for full-year 2025, and officially announced the M6, the share price showed no sign of a meaningful rebound.

Amid this mounting unease, Seres has offered an answer—one that comes in the form of a new company with a symbolic name: “Phoenix.”

In January 2026, corporate registry data quietly revealed the establishment of Shanghai Seres Phoenix Intelligent Technology Co., Ltd., with registered capital of RMB 50 million and full ownership by the Seres Group. Its stated business scope cuts straight into the heart of artificial intelligence: intelligent robotics R&D, foundational AI software, and theoretical and algorithmic AI development.

Once crowned by the capital markets as the “flagship beneficiary of Huawei ADS sentiment,” Seres now finds itself at a precarious crossroads. This is not a short-term fluctuation, but a collective anxiety triggered by a weakening of the underlying narrative. Investors are recalculating the true value of being a “Huawei partner.”

The temperature gap widens as dividends fade

Rewind two years, and Huawei flagship stores were packed with customers drawn almost exclusively by AITO. At the time, AITO was the only four-wheel intelligent terminal in Huawei’s showrooms, enjoying unparalleled exposure and attention.

Today, the center of gravity has shifted. Zhijie and Xiangjie now occupy the most prominent positions in Huawei stores. Feedback from sales staff at core locations in Beijing and Shanghai suggests that many customers now look past the AITO M9, gravitating instead toward the more executive-styled Zunjie, the aggressively designed Xiangjie, or the more value-oriented Zhijie.

Has Seres made a mistake? Not at all. AITO remains a solid product, and the Chongqing super factory continues to deliver at efficiency levels that surpass many new-energy peers. Yet this highlights the harshest truth in business: aesthetic fatigue and dividend dilution.

In the overlapping worlds of consumer electronics and intelligent vehicles, novelty itself is a premium. As Huawei’s “technology shelf” opens more evenly to four partners—Seres, Chery, BAIC, and JAC—and as brand momentum increasingly tilts toward Zunjie, Seres transitions from being the sole favorite to merely the eldest son.

The eldest son is steady and dependable, but also the easiest to overlook. Within the same ecosystem, the market now sees alternatives that feel more distinctive and more imaginative.

For Seres, this means not only a potential diversion of sales, but a downward shift in its valuation anchor. Where the market once assigned Seres a “technology stock” multiple for being the exclusive carrier of Huawei’s intelligent driving, it is now leaning toward a manufacturing valuation as Huawei’s circle expands. Few things are more damaging to a company than this kind of narrative re-rating.

When HarmonyOS traffic no longer favors a single brand, Seres’ moment of real danger begins to emerge.

From localized risk to systemic fragility

A single-product strategy delivers extreme efficiency in favorable conditions, but extreme vulnerability when the wind turns.

For companies embedded in a giant’s ecosystem, even a minor strategic adjustment by the giant can translate into systemic risk. A subtle change in HarmonyOS traffic allocation, or a few extra mentions of Zunjie at a product launch, can erase tens of billions in market value from Seres overnight.

Looking back, Huawei’s original choice of Seres was an asymmetric game. What Huawei valued was not Seres’ brand premium, but its flexibility, its strong supply-chain base, and—most critically—its willingness to fully align. This partnership fueled Seres’ early surge and made it one of the most successful transformation stories in China’s automotive industry.

Yet every gift comes with a price.

This deep binding left Seres without an independent moat. As the “Huawei concept” becomes diluted and more automakers gain access through Hi or selective partnership models, Seres’ scarcity disappears. The core paradox is clear: Seres will not lose Huawei, but it is losing the market’s exclusive attention to the Huawei narrative.

“Without Huawei, there would be no AITO today,” said Seres Group Vice President Kang Bo in response to calls for early independence from Huawei. The more successful Seres becomes, the harder its future becomes. Huawei’s empowerment is evident in AITO’s remarkable sales, and replicating that path across more partners is simply more efficient for Huawei.

Huawei will never all-in on Seres. But Seres must remain all-about Huawei.

Phoenix takes flight: a defensive offensive

Against this backdrop, Seres has no choice but to move fast.

The establishment of Shanghai Seres Phoenix Technology Co., Ltd. is a revealing signal. With RMB 50 million in capital and full ownership, its mission goes far beyond vehicle manufacturing, explicitly targeting intelligent robotics and AI software.

The name “Phoenix” is deliberate. It reflects a desire to rise from the ashes, to evolve from a mere vehicle assembler into a creator of intelligent entities.

Equally telling is its location. Phoenix is based in Shanghai, not Chongqing, signaling a deliberate pursuit of top-tier AI talent and capital operations. It is designed as a new organism, separate from the existing automotive system.

The symbolism deepens as Seres simultaneously strengthens ties with Volcano Engine. The fire of the phoenix may well be volcanic.

Why would Seres seek deeper collaboration with ByteDance after already committing so fully to Huawei? To some investors, this appears counterintuitive, even disloyal. From management’s perspective—watching a steadily weakening share price late at night—it makes perfect sense. This is a defensive offensive, a calculated bet on a second computational anchor.

Seres remains committed to Huawei’s ADS and HarmonyOS cockpit, but in embodied intelligence and broader AI applications, it no longer wants to be just an executor. What it values in Volcano Engine is ByteDance’s core strength: recommendation algorithms and large generative models.

Autonomous driving depends on perception and control. Humanoid robotics, by contrast, demands generalized decision-making and human-machine interaction—areas where ByteDance holds unique advantages.

This is leverage upon leverage. By introducing Volcano Engine, Seres does not abandon Huawei, but injects another super variable into its capital narrative. The downside is equally clear: the “pure HarmonyOS” label fades, and a return to being a contract manufacturer for tech giants becomes visible on the horizon.

The ultimate wager: from cars to intelligent agents

Without this move, Seres’ valuation would remain permanently capped by a manufacturing ceiling. Its price-earnings ratio would swing violently with AITO sales before settling into the 10–15x range typical of conventional automakers—an outcome Seres cannot accept.

The future Seres is likely to pursue a dual-engine strategy. One hand remains firmly on Huawei, using AITO to sustain cash flow and profits. The other bets aggressively, through Phoenix and Volcano, on embodied intelligence and humanoid robotics.

The risks are enormous. Robotics has a far longer commercialization cycle than automobiles, with capital demands that feel bottomless. Yet in 2026, Seres appears to have no alternative.

When danger closes in, yesterday’s shelter becomes tomorrow’s grave. Seres understands this. It must move, and it must change. This may be its most dangerous moment—but also the beginning of true independence. In capital markets, nothing is more frightening than the loss of imagination.

Even if this path looks like divergence, survival comes first. All loyalty exists to ensure continuity. For investors, understanding Seres’ anxiety may matter more than understanding its financials.