Qianwen is about to go head-to-head with WeChat on Agent.

The capital market has not been so excited about a product rumor for a long time.

On June 2, 2026, news emerged that Tencent was preparing to embed Agent capabilities within the WeChat ecosystem. Even though the news had not been officially confirmed, the market was already reacting. On that day, Tencent's share price surged more than 10% in a single day.

Previously criticized for slow moves in AI and a prolonged decline, Tencent finally regained some attention from investors.

A broker analyst focused on the TMT field told Wallstreetcn, "The market is scrambling to buy, because Tencent is about to add a layer of agency that can get things done, on top of the social relationship chain with over a billion daily active users in WeChat. The imagination could support a round of re-pricing."

But as WeChat Agent sparks heated discussion, Alibaba is also quietly preparing with full force.

Wallstreetcn has learned that Alibaba is accelerating the pace of opening up the Agent ecosystem. Recently, Qianwen has announced its entry centered around high-frequency consumption scenarios such as travel and dining. Companies such as China Eastern Airlines, KFC, and Luckin Coffee have begun participating in Agent cooperation tests. Meanwhile, Alibaba businesses like Fliggy and Taobao are also gradually integrating Agent capabilities.

Alibaba brings a familiar story, preparing to redo the matchmaking of supply it has mastered over the past twenty years using Agents, and to compete head-to-head with Tencent.

A competition among top players

The WeChat Agent is getting closer to launch.

A person close to the WeChat Agent project revealed that the original plan was to start internal testing in early June, but the WeChat team is considering further refinement and optimization, so the launch might be postponed. According to the Financial Times, there may be several internal test versions of WeChat Agent.

Outside, Alibaba is quietly making moves, preparing to open the gates to its social circle.

In the past half year, Alibaba has been stress-testing its own ecosystem. Dozens of Alibaba Agents for travel, shopping, government services, etc., have gradually been connected. AI shopping Agents have been embedded in Taobao, Flash Sale is testing instant retail Agents, Fliggy is running travel planning Agents, and Xianyu is also testing.

An industry practitioner told Wallstreetcn, "Actually, the largest AI intelligent trading commercial terminal in China is Qianwen." But he added, "The supply side’s supply form can still be richer."

Zhu Lin’s words proved prophetic; just before WeChat Agent debuted, Alibaba made the first move.

On June 3, Qianwen announced it would fully open up to third-party Agents and Skills — enterprises can operate their own brand Agents on Qianwen. Luckin Coffee, KFC, Mixue Ice City, China Eastern Airlines, and others are testing Agent services on Qianwen and will gradually go online.

Self-prototyping and then expanding the platform ecosystem is Alibaba's usual strategy.

A senior executive within Alibaba explained to Wallstreetcn: "The essence of Agents is to improve service quality, solve customer issues, and enhance the efficiency of customer service, ultimately winning consumers to choose this service."

"For example, if you use Qianwen to hail a ride, just tell it via voice where you want to go and specify a price between 10-15 yuan, and the options will come up immediately." The person said, "No need to open an app, wait for the input box, or fill out forms."

In their view, Alibaba encourages players to create more Agents. "Now, we need to focus on getting Agentic commerce to move within China's large economic environment. Otherwise, all commerce can only write external coding."

Wu Jia, president of Qianwen Business Group, told Wallstreetcn, "What we really want is to let AI blend into ordinary people's everyday life scenarios." He firmly believes this will definitely happen in the future.

Right now, putting Tencent and Alibaba side by side, the most fundamental difference is how users will use this new product.

Users in WeChat mainly seek social interaction, communication, and relationship maintenance. All the utility features (payment, mini-programs, official accounts) are byproducts of social scenarios.

Users in the Qianwen App have more pure purposes: getting things done. Booking flights, ordering coffee, checking points, planning itineraries — they come to Qianwen not to waste time, but to finish a specific task and leave.

With this mindset, the Agent doesn’t need to compete for social attention, just needs to get the task done smoothly. This difference will shape the future divergence in their business models.

Tencent's challenge: Adding a new role to an already dense social space

Connecting the Agent to WeChat sounds like going with the flow, but it may not be as simple as imagined.

The product experience density of WeChat is top-tier even among global internet products. A single chat window carries instant messaging, the Moments entry, official account subscription, video channel, mini-programs, payments, enterprise communication, government services... Any new feature added squeezes out established user muscle-memory interaction paths.

Moreover, to a large extent, WeChat’s core mentality is "talking to people," not "talking to systems." When users open WeChat, the default on the other end is a real person — friend, colleague, family, or someone familiar from a group.

This mindset is so deeply anchored that even the presence of mini-programs is suppressed; most users only open mini-programs when pulled in by a specific scenario, then leave after use.

To put an Agent in such a product, Tencent has to answer: Why would users want to turn from chatting with friends in the same window, to saying to an AI “help me book a ticket”?

Of course, Tencent could sidestep this by not touching WeChat’s main chat framework and instead creating an independent AI entry. But WeChat Agent would then just be another entry, unable to benefit from the real traffic dividends of WeChat; any change must bear the risk of breaking user habits.

This is what Tencent truly needs to prove: In an already mature and highly saturated national-level product, can it create a new AI-native mainstream interaction? The challenge is not technology, but product discipline — what to move, what not to move, and to what degree.

Where WeChat Agent will ultimately go is currently uncertain.

It could be a social + AI hybrid, or skew more towards utility. Whatever the choice, Tencent must face the habitual usage of its 1.4 billion users, which is both its greatest asset and its burden.

But looking back, it’s not as if Tencent hasn't done something similar. WeChat Pay was forcibly carved out as a financial scenario within social spaces, leveraging the unique tool of red envelopes. Video Channel entered the mature short video track by cutting in through WeChat's relationship chain.

Tencent succeeded in both, but each took three to five years. For this Agent battle, the market is willing to pay a premium of HK$410 billion up front, and Tencent will need to deliver real results in the future.

Alibaba's challenge: Building Agent commercial infrastructure

For Alibaba, it doesn’t own a national-level chat product like WeChat. This starting point means Alibaba cannot take Tencent's path of "adding a layer of AI in the familiar social space."

But Alibaba has its own trump card.

From the perspective of industry insiders, whether an Agent can help users book flights depends on whether airlines are willing to open their systems — flights, seats, change rules, meal preferences — for Agents to call, and whether they're willing to allocate resources for this operation chain. The same goes for coffee brands, restaurants, and government services.

Without this infrastructure, Agent business cannot turn; large model companies aren’t good at collecting payment, merchants don't trust decisions made by Agents for users, users are reluctant to hand over payment authority.

This is a comprehensive competition of ToB sales networks, merchant operation systems, and supply chain integration capabilities.

Alibaba has accumulated experience in these areas for over twenty years — from connecting trading factories and overseas buyers during the B2B era, to connecting shops and consumers in the Taobao era, and connecting restaurant brands and takeaway users in the local services era.

This intrinsic gene is not something other large model companies can grow from scratch.

Not only is the foundation important, Alibaba has also taken the lead in thinking about demand in the Agent era.

A person close to Alibaba gave an example to Wallstreetcn: If a user thinks a product worth 10,000 yuan is only worth buying at 7,000, this demand was almost invisible to merchants before. But Agents can keep the user's request open long-term: "Notify me at 7,000 to buy."

As more users express such needs, merchants see real buying intent instead of vague traffic. This is a bit like reverse group-buying, and Agents could even help merchants discover demand.

Merchants previously didn’t know users' real psychological prices. In the future, Agents may directly feed back demand to the supply side, so the business world will have nearly real-time demand signals.

In the Agent era, mastering intent will possess the greatest value, and Alibaba is linking merchant Agents and consumer Agents together.

Looking back now, Alibaba's direction and path are quite clear.

Over the past year, Qianwen has continuously invested in AI shopping, flash sales and other businesses to serve as showrooms and stress tests. After these chains run smoothly, they will be open to all merchants.

From this angle, Alibaba is building the underlying network for the Agent era. If this story is successful, Alibaba and Qianwen could very well see a new round of re-evaluation.

Qianwen’s current position is a bit like mini-programs in 2017. The market previously regarded it more as Alibaba's investment in AI infrastructure—in other words, a "money-burning project."

But when daily service conversations exceed 100 million, 130 million users are using Agents for tasks, and 90% of 300 million transactions come from Qianwen, it already meets some platform valuation criteria.

Even more noteworthy is how the logic of traffic value changes.

Once Qianwen opens to third-party Agent integration, it will no longer just be an internal tool for Alibaba, but a B+C two-sided platform. This gateway could become a new channel for Alibaba’s local life and e-commerce businesses to acquire incremental users, and may also reshape the business models of Taobao, Fliggy, and Flash Sale respectively.

Of course, this path also has its uncertainties: the timetable for Agents to reach large-scale application, whether Alibaba can maintain its pace if Tencent uses WeChat to counter-position.

But one thing is certain: the Agent battle will not be a winner-takes-all story, because Doubao, Qianwen, and ChatGPT are fundamentally on different tracks.

Doubao relies on Byte's traffic genes to perfect companionship and content; ChatGPT relies on OpenAI's enterprise customers to sell tools; Qianwen relies on Alibaba's commercial network to build a platform. Ultimately, these three paths will occupy different ends of the AI application track.

It’s just that, when an Agent can remember your preferences, proactively remind you to order before the lunch peak, help you figure out the optimal redemption plan when your points are about to expire, and place an order as soon as your desired price appears, you’ll find that the past twenty years’ "use and go" user relationships on the internet are being replaced by an AI assistant that understands you.

And this replacement is highly likely to unfold in the home courts of Alibaba and Tencent, simultaneously in two completely different ways.

Risk warning and disclaimerThe market has risks, investment must be cautious. This article does not constitute personal investment advice and does not take into account individual users' special investment targets, financial conditions or needs. Users should consider whether any opinions, views or conclusions in this article suit their particular circumstances. Invest accordingly at your own risk.