Ma Yun oversees the campaign, with "Thousand Questions Sword" targeting the high ground of the "AI gateway".

```

“I looked up and unexpectedly saw Mr. Ma.”

On the evening of February 4, an Alibaba employee posted on Xiaohongshu, sharing a candid photo in which Jack Ma, accompanied by Alibaba partner Shao Xiaofeng, visited the Qianwen project team’s office area. This inspection came just before Qianwen launched its 3-billion-yuan “Spring Festival Treat Plan.”

Two days later, on February 6, the largest red envelope battle among AI apps commenced. Qianwen launched free milk tea orders, leading to a traffic surge: the Qianwen app crashed for a short time, order volumes for food delivery skyrocketed, and consumers reported extended delivery times. Many milk tea stores reportedly stopped taking new orders due to overwhelming demand.

Qianwen crashed, Alibaba went viral, and, with the “milk tea offensive,” the top spot on the App Store unsurprisingly changed hands: Alibaba’s Qianwen app shot to number one, pushing Tencent’s Yuanbao app—which had given away 1 billion yuan in red envelopes just days earlier—into second place.

But this was only the beginning. This wave from Alibaba’s Qianwen pulled in other group businesses such as Taobao Flash Sale, Fliggy, Damai, Hema, and Tmall Supermarket, all preparing to join the Spring Festival festivities with free orders across China. This is the largest investment Alibaba has ever made in a Spring Festival campaign. Among officially announced internet giant AI activities, Qianwen has the largest investment.

To seize the “AI entry point,” domestic internet giants ByteDance, Tencent, Alibaba, and Baidu have all thrown huge sums of money, each hoping to stage a “Pearl Harbor attack” during the Spring Festival’s traffic peak. Alibaba’s strategy is to connect its ecosystem of businesses, using the momentum of lavish cash giveaways to strengthen Qianwen as the Group’s super entry portal.

But this is something the capital markets—especially overseas investors—cannot understand. Even as large language models show remarkable capabilities, the underlying logic of internet growth has not changed: burning money and viral campaigns are still the crude but most effective way to acquire users. As a result, since the announcement of the red envelope war, Alibaba, Baidu, and Tencent shares have each fallen for several days.

Is this old-school internet strategy still effective in the new AI world?

Real-Time, Wartime

On the eve of the 2026 Spring Festival, the air in Xixi Park, Hangzhou, is filled not only with the excitement of the new year but also the smoky gunpowder of an “internet wild era.”

Earlier this week, Jack Ma appeared at Alibaba’s headquarters, walking straight to the “Qianwen Spring Festival Project Team” office. In photos from the scene, Ma wears casual clothes and his signature infectious smile; behind him, a sign inspired by the Year of the Horse reads “Qianwen C-end Business Group.”

In recent years, although Jack Ma has been “retired,” every public appearance has hinted at Alibaba’s strategic direction. From his early fascination with agricultural technology to personally endorsing AI projects now, this shift in focus is itself a rallying call.

According to insiders, Ma was focused on Qianwen’s Spring Festival “user engagement.” He knows that in this wave of tech transformation, failing to control the C-end (consumer) user mindshare would relegate all technical achievements to mere B-end (business) tools.

Industry insiders believe that though Alibaba’s empire is vast, for a long time each business line operated independently, like warlords. To have Qianwen—just one app—seamlessly mobilize resources from other business lines is no small feat.

If Qianwen does become “the strongest human AI assistant,” Alibaba will leap from “shelf e-commerce” to “command e-commerce.”

This would not only fix Alibaba’s lack of a high-frequency C-end traffic portal but also create a closed loop with itself at the core. Users will ask for directions, hail rides, eat, and shop all within Qianwen—keeping all the data and transactions in the Alibaba system instead of relying on external traffic.

And so, the trumpet sounded.

On February 6, Alibaba Qianwen announced a 3-billion-yuan spend to launch the “Spring Festival Treat Plan,” with full-scene free orders covering food, fun, travel, and shopping. In addition to direct subsidies, Qianwen was the exclusive title sponsor of the Spring Festival galas for four major satellite TV channels: Dragon TV, Zhejiang TV, Jiangsu TV, and Henan TV. AI-generated video, image recognition, and Q&A were featured in the gala’s programs and interactive sessions.

During the galas, Qianwen issued exclusive code-word red envelopes, allowing viewers to grab red envelopes in the app while watching. Alibaba appears determined to win this Spring Festival AI war. It brings to mind the famous “ride-hailing war” and “Pearl Harbor red envelope attack” from a decade ago.

In fact, the Spring Festival has always been the proving ground for national-level habit formation in Chinese internet history. Many apps have rocketed to fame thanks to Spring Festival marketing. For Alibaba, subsidies are something like muscle memory. In the past, Alibaba’s Spring Festival campaigns focused on “collecting luck” games and e-commerce red envelopes, with the main goal of promoting mobile payments and boosting e-commerce sales.

Alibaba is making a high-profile move at this time, clearly hoping to use the Spring Festival—the peak period of population movement and information exchange in China—to recreate the legendary “Five Fortunes” campaign by Alipay, and make “Got a question? Ask Qianwen” as commonplace as “scan here” among Chinese users.

Judging from investment size, Alibaba’s 3-billion-yuan initiative is the largest this year or in any previous year, easily surpassing other internet giants. Tencent put up 1 billion yuan; Baidu 500 million yuan; both far less than Alibaba.

But Alibaba is not the only one storming the Spring Festival beaches.

Clash of the Tech Giants

Five days ago, Tencent launched a Spring Festival red envelope campaign on the Yuanbao app, letting users share 1 billion yuan, with single prizes up to 10,000 yuan.

For the past year, Tencent has been relatively “Buddhist” about AI; compared to Baidu’s Wenxin, Alibaba’s Qianwen, and ByteDance’s Doubao, Tencent seemed a step behind in market discussions. CEO Pony Ma hopes that this “cash giveaway” will recreate WeChat’s red envelope moment from 11 years ago.

Meanwhile, Baidu’s Wenxin Yiyan is on the move, preparing to attack during the Spring Festival gala. On January 25, Baidu announced it would give away 500 million yuan for the Year of the Horse gala; ByteDance is focusing on the gala as well, planning a national awareness campaign through sponsorships, interactive features, and AI-powered performances.

It turns out that even with today’s saturated traffic era, financial incentives are still effective during Spring Festival.

With massive subsidies, Tencent’s Yuanbao climbed to the top of the App Store’s free chart on February 1—the first time since its launch. On February 6, Qianwen came in and changed the game. According to Alibaba sources, Qianwen received over 5 million orders within five hours after the event started. Qianwen quickly seized the top spot in all major app stores.

But subsidies are never the end goal—they’re just tangible signs of corporate anxiety.

A source close to the big companies told Wallstreetcn: “For ByteDance, Tencent, and Alibaba, what really matters today is not intelligence per se. Doubao, Yuanbao all care very much about their DAUs; Alibaba is betting on Qianwen, Quark, and Afu. They’re spending heavily for this.”

In the past year, China’s AI large model market saw intense chases and rapid iteration, but the market landscape changed every time the models upgraded. Under these conditions, the key to differentiation might no longer be the models themselves, but rather who can steadily reach users. Thus, big companies are pouring resources into the same direction—seizing the high ground of “entry points.”

“Experience from the mobile internet age has repeatedly shown: once an entry point is solidified, latecomers have a hard time breaking through even with great products,” explained a tech employee.

“If there’s a new traffic entry point, missing it is unthinkable for them,” the source added.

For more than ten years, China’s mobile internet has been app-centered, with platforms controlling distribution, advertising, and monetization. But the rise of AI agents is changing this structure. Once AI can understand intent, decompose tasks, and perform auto operations, users may not need to enter a particular app at all.

So, no platform company can afford to take the AI entry-point battle lightly. Seemingly a must-win blitz war, but the capital market reaction has been chillingly cold.

The Trillion-Yuan Dilemma

Soon after news broke of Jack Ma’s visit and Qianwen’s 3-billion-yuan subsidy plan, the share prices of Alibaba and Tencent fell rather than rose.

Behind this “good news becomes bad news” phenomenon is investors’ concern over the giants’ “path dependency.”

An early-stage market investor told Wallstreetcn: “Big tech is simply repeating what was done in the last decade of mobile internet. At heart, it’s still a battle for traffic.”

In the mobile internet era, Meituan and Didi could burn money to win because the marginal cost of running these services was near zero. Once user habits formed, the cost of serving another user was negligible, unleashing scale benefits and huge profits. However, the economics of generative AI are entirely different. Every AI conversation and reasoning incurs real computing and power costs.

If 3 billion yuan in subsidies brings in 100 million daily active users, it marks not the end of the battle but the start of losses. The more active users, the higher the GPU server bills. If monetization can’t outpace the surge in AI inference costs, more users simply means more pressure.

But most investors care about when they’ll get their money back from this traffic war. Falling stock prices are rational capital voting with their feet—they fear the giants will fall into an arms race of “revenue without profit,” or even “more revenue, worse losses.”

Also, cash-burning subsidies are just the start: they may create short-term miracles but don’t answer the ultimate question—will users stay?

Jack Ma once changed how Chinese people paid because payment tools have high switching costs and network effects. But AI chatbots are different: users’ loyalty is low. They are only loyal to a “smarter brain.”

Wu Jia, president of Qianwen C-end business group, told Wallstreetcn: only those who build truly intelligent models and invest heavily in their ecosystem will reach the final table; such companies are rare.

If next year’s GPT-6 or any other model trounces the current version, users will abandon the red-envelope-giving app without hesitation. What subsidies buy may just be “coupon chasers,” not true heavy AI users.

The real AI revolution lies in a qualitative leap in productivity—AI solving complex industrial issues—rather than simply grabbing consumer attention with cash giveaways.

Recent stock declines may be the market warning these giants: the old internet map won’t reveal the new world of the AI era.

Thirty billion yuan in Spring Festival giveaways may buy a week of buzz, but not a moat for the AI age.

For Jack Ma and his Alibaba, the real challenge of this battle is not how many red envelopes can be given out tomorrow, but after the Spring Festival, once the subsidies fade, how much true user value will remain in that costly chat box.

Risk Warning and DisclaimerThe market carries risks; investment needs caution. This article does not constitute personal investment advice nor consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, views, or conclusions in this article meet their own circumstances. Invest accordingly at your own risk. ```