Did Alibaba simply kneecap its highly effective Qwen AI group? Key figures depart in wake of newest open supply launch

Alibaba's Qwen group of AI researchers have been among the many most prolific and well-regarded by worldwide machine studying group — delivery dozens of highly effective generalized and specialised generative fashions beginning final summer season, most of them solely open supply and free.

However now, simply 24 hours after delivery the open supply Qwen3.5 small mannequin collection—a launch that drew public reward from Elon Musk for its "spectacular intelligence density"—the undertaking’s technical architect and a number of other different Qwen group members have exited the corporate below unclear circumstances, elevating questions and considerations from around the globe in regards to the future path of the Qwen group and its concentrate on open supply.

The departure of Junyang "Justin" Lin, the technical lead who steered Qwen from a nascent lab undertaking to a worldwide powerhouse with over 600 million downloads, alongside two fellow colleagues — employees analysis scientist Binyuan Hui and intern Kaixin Li — marks a risky inflection level for Alibaba Cloud and its function as a global open supply AI chief.

These three Qwen Group members introduced their departures on X right this moment, although they didn’t share the explanations or whether or not or not it they have been voluntary. VentureBeat reached out to sources at Alibaba for extra data and can replace once we receive it. Lin himself signed off with a easy publish: "me stepping down. bye my beloved qwen."

Whereas the corporate celebrates a technical triumph, the sudden exit of its core management suggests a deepening rift between the researchers who constructed the fashions and a company hierarchy now pivoting towards aggressive monetization.

The departing researchers' ultimate present: pocket-sized intelligence

The Qwen3.5 small mannequin collection (starting from 0.8B to 9B parameters) represents a ultimate masterstroke in "intelligence density" from the founding group.

The fashions make use of a Gated DeltaNet hybrid structure that permits a 9B-parameter mannequin to rival the reasoning capabilities of a lot bigger programs.

By using a 3:1 ratio of linear consideration to full consideration, the fashions keep an enormous 262,000-token context window whereas remaining environment friendly sufficient to run natively on commonplace laptops and smartphones — even in net browsers.

Lin, a PKU humanities graduate and polyglot, has lengthy advocated for this "algorithm-hardware co-design" to bypass compute constraints—a philosophy he detailed on the January 2026 Tsinghua AI Summit.

For the developer group, Qwen3.5 wasn't simply one other replace; it was a blueprint for the "Agentic Inflection," the place fashions shift from being chatbots to autonomous "all-in-one AI employees" able to navigating UIs and executing complicated code.

The enterprise dilemma

For the 90,000+ enterprises presently deploying Qwen by way of DingTalk or Alibaba Cloud, the management vacuum creates a disaster of confidence.

Many firms migrated to Qwen as a result of it provided a "third means": the efficiency of a proprietary US mannequin with the transparency of open weights.

Alibaba has not too long ago consolidated its AI efforts into the "Qwen C-end Enterprise Group," merging its mannequin labs with shopper {hardware} groups. The objective is obvious: transition Qwen from a analysis undertaking into the working system for a brand new period of AI-integrated glasses and rings.

Nevertheless, the reported appointment of Hao Zhou, a veteran of Google DeepMind’s Gemini group, to guide the Qwen group signifies a shift from "research-first" to "metric-driven" management.

Trade analysts, together with these cited by InfoWorld, warn that as Alibaba pushes to fulfill investor calls for for income progress, the "open" in Qwen’s open-weight fashions could grow to be a secondary precedence — just like what we noticed with Meta after the disappointing launch of its Llama 4 AI mannequin final spring, and subsequent reorganization of its AI division, seeing the hiring of Scale AI co-founder and CEO Alexandr Wang and following departure of preeminent researcher Yann LeCun.

Enterprises counting on the Apache 2.0-licensed Qwen fashions now face the chance that future flagships —such because the rumored Qwen3.5-Max—can be locked behind paid, proprietary APIs to drive Cloud DAU (Day by day Energetic Consumer) metrics.

The takeaway? If you happen to worth Qwen's open supply efforts, obtain and protect the fashions now, when you nonetheless can.

The "Gemini-fication" of Qwen?

The inner friction at Alibaba mirrors the tensions seen at OpenAI and Google: the "soul" of the machine is usually at odds with the "scale" of the enterprise. Xinyu Yang, a researcher at rival Chinese language AI lab DeepSeek, captured this sentiment in a stark publish on X: "Change the superb chief with a non-core individuals from Google Gemini, pushed by DAU metrics. If you happen to choose basis mannequin groups like shopper apps, don’t be shocked when the innovation curve flattens."

This "Gemini-fication"—the shift towards a extremely regulated, product-centric tradition—threatens the very agility that allowed Qwen to surpass Meta’s Llama in by-product mannequin creation. For the worldwide AI group, the lack of Junyang Lin is symbolic.

He was the first bridge between China’s deep engineering expertise and the Western open-source ecosystem. With out his advocacy, there are fears that the undertaking will retreat right into a "walled backyard" technique just like its Western rivals.

'Leaving wasn't your alternative'

The technical brilliance of the Qwen3.5 launch has been overshadowed by the heartbreak of its creators. On social media, the sentiment among the many group members who constructed the mannequin is one among mourning fairly than celebration:

Chen Cheng, a Qwen contributor, explicitly alluded to a compelled departure, writing in a publish on X: "I'm actually heartbroken. I do know leaving wasn't your alternative… I truthfully can't think about Qwen with out you."

Li prompt the exit signaled the top of broader ambitions, comparable to a deliberate Singapore-based analysis hub: "Qwen may have had a Singapore base, all because of Junyang. However now that he's gone, there's no cause left to remain right here."

What occurs to Qwen's open supply AI efforts from right here on out?

The recognized details are easy: Qwen has by no means been technically stronger, but its founding core has been dismantled. As Alibaba prepares to face buyers for its fiscal Q3 earnings report on March 5, the narrative will doubtless concentrate on "effectivity" and "business scale."

For the enterprises presently excited in regards to the 60% value reductions promised by Qwen3.5, the rapid future is vivid.

However for the bigger AI group, the price of that effectivity stands out as the lack of probably the most vibrant open-source lab within the East.

As Hao Zhou takes the reins, the world is watching to see if Qwen stays a "mannequin for the world" or turns into merely a part in Alibaba’s company backside line.