Alibaba is fighting a hard battle that will determine its fate

In September 2025, Alibaba's core decision-making team held a series of meetings with senior executives from Quark and other businesses that would "impact Alibaba's future destiny."

During this series of meetings, the group's core decision-making team made a decision: to fully promote a C-end AI project named "Qianwen APP," viewing "Qianwen APP" as the core of Alibaba's entire C-end AI market layout—Qianwen APP will directly enter the C-end market; in addition, the AI C-end products already released by Alibaba and other upcoming AI C-end products will also revolve around "Qianwen APP" and the underlying "Tongyi Qianwen" large model.

This means that Quark Cloud Disk, browser, the upcoming AI glasses, and other C-end AI scenarios that Alibaba will attempt will all be integrated into the Qianwen ecosystem. In other words, Qianwen is expected to "become the mental symbol of Alibaba AI C-end, similar to how Taobao is for Alibaba's e-commerce."

In this series of meetings, some executives described "Qianwen APP" as "a key battle that will influence Alibaba's future." Insiders revealed that a very high-ranking group figure explicitly stated that "Qianwen APP" is not only a strategic project that will affect Alibaba's future destiny but also an "overall battle" that requires collaboration from all departments across the group. Subsequently, on a Wednesday at the end of October, a kickoff meeting was organized with participation from core internal personnel.

96 hours after the kickoff meeting was a Monday.

During these 96 hours, over 500 engineers, gathered from Beijing, Guangzhou, and Hangzhou, had secretly moved into the third and fourth floors of Building C4 at Alibaba's Xixi Park in Hangzhou. They were almost the "elite generation" of Alibaba's internal AI products at that time: there were post-95s prodigies who had been carefully nurtured by Alibaba since graduation, AI academic experts recruited from overseas with significant investment, senior product managers who had participated in multiple product launches and tough battles, and specialists with expertise in subfields like multimodal…

The workstations, accommodations, and project divisions for over 500 people were quickly "settled" within 96 hours, and specially printed badges were handed to them after they moved into the C4 secret work area on Monday. Most of them did not know they were joining a "top secret" strategic project within Alibaba until they boarded the flight to Hangzhou; they were only informed that the company urgently needed their participation in a key project.

In the following week, the number of elite personnel parachuted in grew to nearly 900, setting a new record for Alibaba in terms of the number of elites drawn from different departments for a single project.

At the end of the fourth floor of Building C4, there was a room named the "War Room"—it was actually a large conference room. Wu Jia, the person in charge of Qianwen APP, moved his office here, surrounded by the workstations of Wu Jia's core subordinates and the heads of various core sections of Qianwen APP Here has become the "battle command center" for Alibaba's tough fight.

There are no OKRs, no rigid frameworks, and not even a comprehensive blueprint called a "roadmap." More than 900 people are closed off on the third and fourth floors of Building C4 in project teams, with data reports and user analyses generated collaboratively by humans and AI delivered to each "frontline soldier" every night; every week, the weekly goals set in the "war room" are refined into specific work plans, broken down to each member of the project team.

Alibaba rarely pushes for the birth of Qianwen in the form of "project teams" on the front lines, rather than through a large organizational structure.

On November 17, more than a month later, the Qianwen APP officially launched. Seven days later, downloads exceeded 10 million. Tiger Sniff learned that this figure significantly surpassed the "expectations" set by the group's core team for the project. Within 10 days of Qianwen's launch, the Quark AI glasses, as a hardware component of the "Qianwen ecosystem," officially went live. Tiger Sniff learned that the total order volume for Quark AI glasses has already far exceeded the previous team's expectations.

From the data, both the Qianwen APP and Quark AI glasses ended their first two battles with small victories. However, it is undeniable that a bigger battle lies ahead for Alibaba: from the perspective of C-end AI products, currently, Doubao, DeepSeek, and Yuanbao firmly occupy the top three spots in the market, while behind them, Kimi and other "six little tigers" are also struggling to survive; in the AI glasses market, Meta, with its first-mover advantage, occupies a vast majority of the global market, while brands like Rokid, Xiaomi, and Yingmu have already secured significant shares in the domestic AI glasses market to solidify their brand recognition.

It is worth noting that these competitors in both the C-end AI APP and AI glasses markets are exceptionally formidable. Taking Doubao as an example, with its "extraordinary" traffic endowment, Doubao emerged from the era of "AI six little tigers dominating the world" in 2024, and despite the successive impacts from DeepSeek and Yuanbao, it has managed to hold its ground, becoming one of the few AI products in the country with "hundred million user volume." In the AI glasses field, Rokid has withstood the impact of giants in the industry with an entrepreneurial spirit and achieved a "mythical record" of "40,000 units sold in 5 days" in the third quarter of this year.

"Qianwen is seen as the era of Eddie Wu, the third critical battle for Alibaba's fate. One is AI + Cloud, one is Taobao Flash Sale, and one is Qianwen," a relevant person told Tiger Sniff. "One of the internal consensus is that if success is achieved in the AI C-end market, Alibaba's potential could multiply several times compared to today."

Fate Decisions Made in Summer

Tiger Sniff exclusively learned that before the summer of 2025, the "Qianwen Plan" had not become the top choice for Alibaba's core decision-making team.

In fact, until the beginning of this year, Quark was still the main character in Alibaba's AI C-end narrative A key factor has become a bargaining chip that influences the direction of decision-making.

In September 2025, Alibaba Cloud released Qwen3-Max, the flagship model of Tongyi's large model with a parameter scale exceeding one trillion. According to Tiger Sniff, when Alibaba's core decision-making team experienced the powerful multimodal capabilities and reasoning research capabilities of Qwen3-Max, their confidence in "developing a pure C-end AI application" significantly increased.

Relevant sources told Tiger Sniff that the core decision-makers, familiar with technology and AI, personally tried most of the C-end AI model tools on the market and compared them with Qwen3-Max. This person believes that Alibaba should quickly adjust its previous thinking and directly enter the C-end market. "We may not have traffic advantages or social advantages, but our greatest confidence lies in our model capabilities. We found that many leading domestic AI applications have actually fine-tuned the Qwen model, so why don't we do it ourselves?"

This sparked a series of "decision meetings" mentioned at the beginning of the article.

From the information gathered by Tiger Sniff, the emergence of the Qianwen APP is not a predetermined strategy from years ago, but a dynamic decision made based on the latest internal and external circumstances.

Once the general direction was determined, the next step was to decide how to proceed.

In the third quarter of 2025, Alibaba's core decision-making team held multiple meetings with the business to discuss how to enter the C-end AI market.

After several communications, the core team gradually formed three basic consensus points, which became the "Longzhong Dui" for Alibaba's AI to C:

1. The Qianwen APP is core, but it is not just an APP; it should penetrate into more scenarios (for example, the Qianwen APP is integrated into Quark AI glasses in the form of an AI assistant and also integrated into browser products in a similar form).

2. Not to create a purely chat companion tool, but to meet users' chat needs while providing "efficiency and professionalism," becoming a productivity tool for users. Considering today's users' needs for chatting, while leveraging Alibaba's model capability advantages. Based on this, the core team provided a six-character guideline: "Can chat, can handle affairs."

3. The present moment (referring to September 2025) is crucial because the Qwen3-Max model capabilities are sufficient to accomplish this task. We must seize this historic window period to create an AI application that can "incorporate the entire Alibaba ecosystem." During this historic window period, investments must be decisive and not hesitant.

After these three basic consensus points emerged, the Qianwen APP project was able to "move forward without obstacles" within Alibaba: the most elite engineers, the most flexible project and management methods, and substantial resource investments followed one after another.

A subsequent question is how to define the boundaries between Quark and Qianwen?

Relevant sources told Tiger Sniff that based on the three basic consensus points, the team has redefined Quark's positioning: Quark will no longer play the "core" role in Alibaba's AI to C initiative, but will focus on its strengths in "search and post-search services (such as cloud storage and scanning)." It is worth noting that, from an organizational perspective, Quark and Qianwen APP both belong to the "Alibaba Intelligent Information Business Group," and Wu Jia serves as the president of this business group while also holding the positions of CEO of Quark and head of Qianwen APP. During the advancement of the Qianwen APP project, a large amount of technical capabilities sourced from Quark have been embedded into the Qianwen APP. One of Wu Jia's key responsibilities is to allocate product capabilities and resources based on the different positioning of the two products.

An Alibaba "Mobilization"

C4 Building, 4th Floor, Xixi Park, Alibaba, Hangzhou.

The room is filled with drinks and snacks, where Wu Jia and core members of Qianwen hold endless review meetings and discussions every day. They often neglect meals, grabbing a bag of snacks to satisfy their hunger "on the spot."

The Qianwen project team is a "special zone" within Alibaba.

One detail is enough to show the uniqueness of this project team; it has broken away from Alibaba's traditional organizational model, management style, and product development system. It is a product born on the "front lines of war." The highest levels of the group have granted this batch of "young commanders on the front lines" great flexibility and space.

Huxiu learned that, to ensure the creativity of the Qianwen project team, the highest levels of the group did not require the team to adopt management methods like OKR. More than 900 engineers who came from various backgrounds have mostly not had specific OKRs for the Qianwen project over the past two months.

Tactical goals are updated weekly based on real data and user feedback, and with the progress of product iterations, everyone clearly understands "the specific output goals for each week."

The budget system is equally flexible. Unlike many large projects within Alibaba, the expense applications for the Qianwen project are assessed weekly, based on the upcoming week's "key milestones, key outputs, and key weekly goals," making the entire budget approval and evaluation process very swift.

The product itself adopts a strategy of "overall Alibaba warfare": the different AI capabilities embedded in the Qianwen APP actually stem from the capabilities accumulated by different business lines in the past.

For example, the multimodal capabilities are handled by elite teams parachuted in from the Tongyi large model and other teams; exam paper recognition and AI medical-related capabilities are managed by elite talents from Quark who excel in these areas.

Each capability segment has established subdivided sub-project teams, and the members of these teams may have previously been distributed across different business segments. Once a capability (function) is completed and enters the testing and launch phase, a separate data analysis team will begin daily data tracking. If the user data (user penetration rate, user feedback) for a certain capability does not meet expectations, the data analysis team will meet with the sub-project team to discuss the underlying logic and then iterate quickly.

A series of user dimension data (DAU, etc.) has become the key "goal" for the entire Qianwen project team, but it is not everything.

Huxiu learned that, due to the "three basic consensus," the Qianwen project is viewed as "not just a simple companion chat AI" but as a "productivity tool AI that can meet chat needs while improving user efficiency and professionalism." Therefore, when the team considers user dimensions, they take into account not only "quantity" but also "quality." There is a separate team that will conduct in-depth research on the characteristics and needs of all Qianwen APP users, exploring how to meet user needs from multiple aspects such as chat companionship and productivity.

In terms of goals and evaluation dimensions, the group's decision-making level has also given Qianwen APP a different "tolerance space" than before. According to Tiger Sniff, until the day Qianwen APP was launched, the decision-making level did not provide the project team with a clear "data indicator red line (i.e., a minimum threshold)," but members of the decision-making level basically became seed users of Qianwen APP. In some separate communication groups, they often share their feedback or suggestions based on their experiences.

From these perspectives, Qianwen APP may bring Alibaba not just an AI C-end product; it actually represents a new "collaboration method" within Alibaba. I can elaborate with an example: these elite members who were transferred to Qianwen actually had clear OKRs in their own business segments before receiving the "transfer notice." After joining the Qianwen project team, their existing OKRs would be in a "pending" state. Tiger Sniff learned that Alibaba had quickly formed a complete set of guarantee plans regarding the performance and treatment of these elites over the past few months. This institutional model has actually created a paradigm for similar "super projects" within Alibaba in the future.

In terms of collaboration, projects like Qianwen that draw on hundreds of elites from dozens of departments do not occur every day within Alibaba. Moreover, as mentioned above, Qianwen actually "adopted various AI capabilities from different departments of Alibaba." Such sincere collaboration of "bringing out the best talents and most valuable technical secrets" across departments can only occur under the premise of the senior management of Alibaba promoting an "overall battle." However, this has clearly set a precedent for future in-depth cross-departmental collaboration within Alibaba.

Conclusion: Just Beginning, A Tough Battle

Looking calmly, the AI to C war triggered by the Qianwen APP has just fired its first shot.

For Alibaba, it is certainly not yet time to celebrate.

Qianwen APP and Quark AI glasses are essentially competing for the C-end AI entry point, which is a battleground for ByteDance, Tencent, Baidu, and Meituan.

When communicating with people related to multiple AI to C products at Alibaba, I asked a key question: How much are you investing in traffic?

From the information I have perceived, they are very clear about the "traffic magic" behind the rise of Doubao, DeepSeek, and Yuanbao over the past two years; they also understand that it is difficult for Alibaba to "wave a blank check to invite guests" in terms of traffic like some competitors.

Relevant individuals told Tiger Sniff that Alibaba's investment in traffic for AI to C is very calm and restrained.

"The internal team is clearly aware that this is not the same as Taobao flash sales; we cannot treat Qianwen the same way we treat instant retail business." This individual believes that internally, there is a clear understanding that instant retail is aimed at the entire C-end market, embracing every C-end user; however, the positioning of Qianwen is that it is not an entertainment product and not merely a chat product "The above-mentioned person said.

But this has led me to think: looking at the fate of several popular C-end AI products in the domestic market over the past two years, they often need to seize the key window period at the very beginning to quickly achieve 'user fission,' then swiftly surpass the milestone of ten million monthly active users, ultimately evolving into a national-level (hundred million user) blockbuster.

The hidden difficulty is that, whether it's Doubao or Yuanbao, they are still in the 'strategy cycle' today, and they still yearn to continue expanding; in the coming months, DeepSeek is likely to release its long-awaited R2. Considering that many of the players mentioned above hold significant traffic themselves, will Alibaba's expansion in the C-end market fall into the narrative of Quark: product capabilities are praised, but marketing and momentum are interpreted as 'slightly inferior'?

Not only that.

According to information obtained by Huxiu, based on Qianwen, Alibaba is attempting to integrate the entire ecosystem, including e-commerce. This is clearly a huge imaginative space. However, to date, including OpenAI, the industry is still in the exploratory stage regarding what AI e-commerce actually looks like. Moreover, the reshaping of e-commerce by AI itself can be divided into two aspects:

· If you have an AI super entrance, then how to drive traffic to e-commerce based on this entrance is what Qianwen aims to do.

· How to use AI to reconstruct existing e-commerce businesses is what Taotian is currently doing.

Alibaba's challenge lies in the fact that these two tasks need to be promoted simultaneously.

This is destined to be a tough battle.

Huxiu

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investing based on this is at your own risk