Companies like Xiaomi, Honor, OPPO, and vivo probably never expected that their biggest competitor would one day not be Apple, but a dominant 'lobster.'

In just a few months, OpenClaw has transformed from a geek toy into a phenomenal trend: Scenes of engineers installing it for free for passersby in front of Tencent's headquarters emerged, and Shenzhen Nanshan Square was once crowded with developers and ordinary people holding laptops in line. Tech giants like Tencent, Alibaba, and ByteDance quickly followed suit, offering one-click deployment templates and cloud images. Derivative tools like QClaw/WorkBuddy have turned WeChat/QQ into direct remote control entry points.

Beyond the 'national lobster-raising' craze, OpenClaw's impact on Mac mini sales and the quiet entry of AI glasses are racing to define the best computing terminal in the AI era. Meanwhile, smartphones—the most popular hardware of the mobile internet era—have fallen unusually silent.

Not only that, but in early 2026, the smartphone market's long-standing dynamics were completely rewritten. This time, price hikes knew no brand boundaries or price exemptions. From Samsung's S26 series flagship phones seeing a uniform price increase of 1,000 yuan to Honor's Power 2 starting price jumping by 700 yuan (a 35% increase), even the usually 'affordable' realme Neo series saw a 300-yuan price hike compared to its predecessor.

Mainstream brands simultaneously launched large-scale price hikes in the same quarter. According to the latest news, OPPO and OnePlus announced price adjustments today, following earlier reports that vivo, Xiaomi, and Honor had all confirmed plans for new price hikes in March, with a possibility of second or even third rounds of hikes in the second half of the year.

The direct trigger for these price hikes was the sharp surge in memory and storage prices. According to TrendForce data, spot prices for smartphone memory chips have risen by over 300% in the past three months. The cost of 12GB LPDDR5X memory skyrocketed from 200 yuan to nearly 600 yuan, while 1TB storage chips similarly jumped from around 200 yuan to about 600 yuan, with 256GB UFS4.0 seeing an 80% to 90% increase.

In the past two years, AI infrastructure construction centered around large models has entered a money-burning race. Tech giants' demand for computing power has exploded exponentially, with model parameter scales quickly advancing from billions to hundreds of billions and even trillions. Multimodal applications like text-to-image and video generation have further pushed the computing power ceiling. Storage manufacturers like Samsung and SK Hynix have swiftly shifted production capacity to higher-margin HBM (High Bandwidth Memory), sharply reducing the supply of consumer-grade DRAM and NAND Flash needed for smartphones. Once supply and demand fall out of balance, price hikes become inevitable.

Thus, a comprehensive price hike spreading from high-end flagships to mid-range and entry-level models has become the most intense price shock the smartphone industry has seen in five years. The hardest hit are mid-to-low-end smartphone brands already squeezed between rising costs and market pressures, with razor-thin profit margins.

The OpenClaw Effect: Will Smartphones Be Abandoned by AI?

To understand today's smartphone industry dilemma, we must first revisit its heyday. Over the past 15 years, mobile internet has built an extremely efficient value closed loop (closed loop): Smartphones are the most private and portable personal terminals, while apps are the most precise service distribution carriers.

Ride-hailing, payments, socializing, shopping—every need was precisely packaged by engineers into an icon, accessible with a single tap. This logic not only created immense commercial value but also shaped a generation of consumers' habit of upgrading every two years. Each new generation of products improved in these key dimensions, and users were happy to upgrade every two to three years to enjoy the leap in experience brought by technological progress.

But AI is fundamentally shaking the traits that smartphones pride themselves on most—portability, ease of use, and always-on connectivity. In the face of AI agents' demands, these traits have become bottlenecks.

From late 2025 to early 2026, an unexpected phenomenon quietly emerged in the tech world: With the explosive popularity of OpenClaw (a powerful AI agent framework), Mac mini sales saw a rare and significant surge. Developers and geeks worldwide scrambled to buy this tiny desktop computer to deploy their own AI agents, even if the functionality was still imperfect, just to experience the future of 'AI automation' firsthand.

In an era where everyone is searching for the 'next AI hardware,' the first to become the best carrier (carrier) for AI was not a flagship smartphone packed with AI features, but a humble desktop computer.

This outcome was no accident. Computers have existed as productivity tools since their inception. Before graphical interfaces, computers only had command lines—black screens with white text and blinking cursors—only accessible to those who could speak 'machine language.'

While graphical interfaces lowered the barrier to entry, the computer's underlying DNA has never changed: a complete file storage system, operable command-line interfaces, and a highly customizable runtime environment. These traits make computers naturally capable of being 'called by machines.' AI now allows ordinary people to issue commands in natural language for machines to execute—and the computer's architecture is naturally prepared for this 'human-orders, machine-executes' model.

Smartphones, on the other hand, followed a completely different evolutionary path. Emerging in the 21st century, their entire interaction logic skipped straight to graphical interfaces, designed to make operations as simple, intuitive, and barrier-free as possible. For ease of use, apps on smartphones were designed as closed, plug-and-play islands, with strictly controlled permissions for each application. The system's core interfaces were largely closed to third parties, and inter-app communication was heavily restricted.

This architecture was an advantage in the mobile internet era but became a fatal constraint in the AI era. It's hard to imagine an app downloaded from the App Store automatically answering takeout calls, checking schedules, or completing shopping, as smartphone apps' permissions and capabilities are almost locked within application sandboxes.

This is why the iPad has never replaced the Mac, and 'smartphone assistant' AI phone solutions are ultimately just workarounds. The answer points to the same place: Computers naturally retain the soil for 'machine interaction,' while smartphones were designed from the start only for 'human interaction.'

Meanwhile, the quiet rise of AI+AR glasses poses an even deeper challenge to smartphones at the 'form factor' level. As AI carriers, glasses require no device retrieval, screen unlocking, or app searching—they're always in your field of view, constantly sensing the environment and responding to needs. This 'app-less' interaction paradigm naturally aligns with AI agents' mode of 'running continuously in the background and proactively serving humans.'

Capital and industry giants' moves best illustrate the certainty of this direction. Meta's Ray-Ban AI glasses have rapidly penetrated global markets. By 2025, two of China's top three internet companies by market cap had clearly entered this track (sector): ByteDance's Pico is about to release AR glasses, and Alibaba's Qianwen AI glasses sold out in three hours.

As previously mentioned in Xin Lichang's articles, when AI becomes powerful enough, the interaction paradigm of 'I open an app' will be replaced by 'AI completes tasks for me.' This shift will progressively but irreversibly disrupt the current smartphone business model.

Faced with the deep anxiety brought by hardware form factor changes, smartphone manufacturers have responded with the most direct strategy: acceleration. Nearly every day in March was filled with new product launches—vivo X300 Ultra, OPPO Find N6, Honor Magic V6 Chinese version, OnePlus 15T, with three or four flagship models debuting in mid-March alone, followed by OPPO Find X9 Ultra in April.

Smartphone manufacturers' strategy of accelerating specs and frequent releases is essentially an 'externalization of internal anxiety.' Before finding a product definition truly suited for the AI era, they maintain consumer attention and upgrade impulse through higher release frequencies and more aggressive configurations.

The problem is that this approach is now encountering both user aesthetic fatigue and shrinking upgrade willingness. With storage prices rising rapidly, launching new products quickly to lock in current cost windows is a more rational choice than waiting. But this creates a paradox: More new products mean less differentiation between them; faster rhythms make it harder to satisfy user expectations for 'truly worth upgrading.'

Data from the China Academy of Information and Communications Technology shows that user upgrade cycles have lengthened from 24 months in 2020 to 33 months in 2025. Under the dual pressure of price hikes and form factor uncertainty, this number continues to grow.

An Unexpected Disaster: Who Killed the Budget Phones?

Today we know that a smartphone's costs are highly concentrated in three core components: The processor accounts for the BOM (bill of materials), followed by memory and storage, and the screen. Together, these three account for over half of the cost. This means price fluctuations in memory and storage directly determine smartphones' profitability.

In 2025, the supply-demand balance in the memory market was completely disrupted, with memory price adjustments not only sharp but also erratic.

What used to be annual or semi-annual price adjustment cycles have compressed to twice a month. For a 12GB+256GB smartphone, overall material costs can fluctuate by over 1,000 yuan within months.

This has caused massive turbulence in the mid-to-low-end market. UBS predicts that by 2026, memory costs for low-end phones will account for 40% to 45% of the BOM, far higher than 20% to 30% in 2025. In other words, for a mass-market phone priced at 1,500 yuan, nearly half of its cost structure will come from memory and storage.

From an economic perspective, this market's contraction represents a reversal of the 'technological deflation' logic of the AI era. For the past 15 years, Moore's Law-driven hardware cost declines made 'affordable high-performance devices' possible. But when AI infrastructure construction demands fundamentally altered the global storage chip supply-demand structure, this logic was interrupted.

Faced with such fundamental changes in cost structure, Xin Lichang believes that mid-to-low-end smartphone brands once known for 'cost-effectiveness' essentially have only three paths:

The first path is direct price hikes. But budget models target the most price-sensitive users, who are extremely sensitive to price changes. Direct hikes would drive many users to the secondhand market, previous-generation flagships, or simply delay upgrades. When Honor's Power 2 rose to 2,499 yuan, public backlash confirmed this market reaction.

The second path is maintaining original prices and absorbing costs internally. This means letting the brand's financial statements bear industry-wide cost shocks. For mid-to-low-end brands relying on volume, higher shipments mean greater losses—a 'sell more, lose more' situation that quickly breaks the brand's cash flow.

Thus, for rational manufacturers, discontinuing production is the most realistic choice. Rather than being trapped on both ends, it's better to proactively cut unprofitable SKUs and concentrate limited resources on higher-margin product lines. IDC predicts that by 2026, the market share for sub-$200 low-end phones will shrink by 4.3 percentage points to 20%. This contraction stems not from consumer willingness but from supply-side retreat.

The market now presents a bizarre sight: Flagship phone launches are bustling, while the budget phone market falls silent. According to media reports, in January 2026, Xiaomi and OPPO cut full-year device orders by over 20%, vivo by nearly 15%, and Transsion to below 70 million units, primarily targeting mid-to-low-end models and overseas products.

The upstream chip market also signals this trend: Counterpoint predicts that in 2026, MediaTek shipments will decline 8% year-on-year, Qualcomm 9%, and UNISOC—which mainly serves the low-end market—will drop 14%.

As budget phones become history, consumers are already voting with their feet. A unique 'reverse replacement' phenomenon has emerged in county-level smartphone markets—with the same budget, consumers now opt for previous-generation flagships on the secondhand market rather than new models, getting better hardware at lower costs. This behavior represents ordinary consumers' most rational yet helpless response to price hikes.

In this industry-wide retreat from the mid-to-low-end market, no one remains unscathed. All brands targeting the lower tiers are experiencing severe pain. First among them may be vivo—the undisputed king of the 'sinking/mid-range market' in recent years.

During the early parameter-driven industry phase, vivo—guided by its internal 'integrity' culture—demonstrated unabashed expansion in the mid-to-low-end price segments. Not only did it launch dense product lines like the Y series, S series, and iQOO to cater to different consumer tiers, but it also built an extremely formidable offline moat. By 2024, vivo had over 250,000 offline stores nationwide (including direct, agent, and franchise locations). This deep offline network vastly increased channel capacity, using astonishing throughput to flood households with vivo's mid-to-low-end models.

According to the latest RD Observation data, vivo dominates the 0-2,000 yuan price segment, holding a 22.0% overall market share (16.8% for vivo, 5.2% for iQOO). Its performance-focused iQOO 15 sold more in 30 minutes than its predecessor did in a full day, with offline sales jumping 280% on launch day—a testament to its massive offline channel presence and core audience loyalty.

Over the past four years, nourished by this sinking market, vivo has repeatedly topped China's smartphone shipment rankings. However, this also proves that its impressive total sales volume heavily depends on the mid-to-low-end market. When AI-driven memory inflation brutally punctured low-end phone profits, vivo's scale-based competitive barrier—once so daunting to rivals—instantly became its heaviest structural burden.

No one feels this pressure more acutely than Hu Baishan. This veteran, promoted to vivo president less than a month ago, hadn't even enjoyed his new role's joys before facing his first major crisis as commander-in-chief.

He inherited a vast channel matrix deeply rooted in the sinking market, once proven unbeatable. But now, the low-margin business logic sustaining this system faces severe external challenges. Leading these 250,000 offline stores—long accustomed to traditional volume-driven models—out of their comfort zones to adapt to this supply chain-induced 'budget phone cyclical retreat' will be the first major obstacle this new president must overcome.

In 1973, when crude oil prices skyrocketed from $3 to $12, the U.S. auto industry attempted to humble itself and learn how to make cheap, fuel-efficient cars. But this humility didn't last long. In 1979, Paul Volcker pushed interest rates to an extreme 18% to combat inflation.

This figure not only shut off the credit tap but also strangled the lifeline of microcars. Small cars like the AMC Gremlin were caught in an awkward financial black hole: they had thin profit margins yet had to bear soaring material costs; they were supposed to be lifesavers but became heavy burdens on financial reports.

When loan interest alone could buy half a car, the middle class began to shift toward a "one-and-done" strategy. Ford and Chrysler executives soon made the most commercially intuitive—yet also the most disheartening—decision: to give up.

Rather than engaging in hand-to-hand combat with Japanese cars in a low-margin red ocean, they retreated behind the defensive walls of high-margin products. They turned to pickup trucks and SUVs—where profit margins were fat enough to cover all costs.

Back then, the collapse of microcars in the U.S. market wasn’t due to technical hurdles. In that turbulent era, Detroit simply chose the "easier money."

Today, Dongguan’s smartphone supply chain is reliving Detroit’s dilemma.

The price hike shows no sign of ending anytime soon.

This wave of price increases isn’t confined to the low-to-mid-end market; high-end flagship models are equally affected. IDC data shows that by 2026, flagship smartphone prices in China will surge by over 30%, with similarly configured models costing 300 to 1,000 yuan more than in 2025, and the price gap for large-storage versions could even reach 2,000 yuan.

High-end Android flagships from Huawei and Samsung are also caught in the storm. But among all brands, Apple faces the most unique situation: it is the least impacted by this price surge and may very well be the only mainstream smartphone maker to maintain its starting prices unchanged in 2026.

Apple’s ability to "stay unscathed" stems from several interlocking structural factors. First, Apple has already locked in procurement prices with Samsung and SK Hynix, accepting long-term agreements that see memory prices nearly double. On the surface, this appears to be a concession by Apple, but in reality, it trades certainty for stability: in a market where prices fluctuate wildly each month, the value of locking in cost curves far outweighs short-term procurement savings.

Second, iPhone models inherently have lower memory capacities than Android flagships (the base iPhone 17 has just 8GB, while the Pro Max has 12GB), so memory accounts for a smaller proportion of the overall BOM, resulting in a smaller absolute financial impact. Apple’s high-capacity versions have long maintained a premium pricing structure: the 512GB model costs 2,000 yuan more than the 256GB version. Thus, even if flash memory costs double, this pricing tier still provides Apple with ample buffer space.

For ordinary consumers, the harshest reality is that this round of price hikes shows no signs of peaking. Ben Wood, chief analyst at CSS Insight, explicitly states that the global shortage of memory chips will persist until 2027, driven by the massive expansion of AI infrastructure—which shows no signs of slowing but is instead accelerating. As long as tech giants continue to race to build larger data centers and train bigger models, the structural squeeze on consumer-grade storage capacity will continue, rather than being a cyclical fluctuation.

Looking further ahead, TSMC’s A14 process (1.4nm-class) is scheduled to enter mass production in the second half of 2028. While more advanced processes bring significant performance/power efficiency improvements, they also come with higher wafer manufacturing costs, which typically push up the cost baseline of flagship SoCs and even high-end smartphones in phases.

From a supply chain perspective, the current cost escalation cycle from 3nm→2nm→1.4nm is expected to persist until 2028–2029. A meaningful easing of cost per performance may not occur until 2029–2031, after the A14 process matures and capacity is fully ramped up.

Until then, consumers and the industry will jointly endure a prolonged period of high-cost adjustment. In other words, smartphone price hikes are unlikely to see substantial relief until after 2029. The intervening period will be a cost-adjustment phase borne by both consumers and the industry.

As smartphones’ best use cases are quietly eroded by computers and glasses, extending replacement cycles and making fuller use of existing devices may be the smartest coping strategy for ordinary users in this industry upheaval. Looking back, every technological paradigm shift has been accompanied by a painful revaluation of old hardware forms.

This time, it’s the smartphone’s turn.

*The featured image and illustrations in the text are sourced from the internet.