Author: Hu Shixin , Deep Web & Tencent News Xiaoman Studio

Edited by: Ye Jinyan

On March 3, Jack Ma, along with Joseph Tsai, Yongming Wu, Xiaofeng Shao, Fan Jiang, and core executives from Alibaba and Ant Group, including Jing Xiandong and Han Xinyi, gathered at Hangzhou Yung Valley School in a rare event. This first meeting of the new year focused on the opportunities and challenges brought by AI, sending a strong strategic signal to the outside world that both groups are "All in AI."

Dramatically, in the early hours of the day after this high-level symposium concluded, Lin Junyang, the technical lead who single-handedly propelled Alibaba's Qwen to the pinnacle of global open source and the group's youngest P10, suddenly posted "me stepping down. bye my beloved qwen" on the social platform X, a post that ignited the global AI community.

Overnight, Alibaba's Qwen lost its "helmsman".

(Lin Junyang X Latest News)

Was Lin Junyang's resignation not voluntary?

Based on multiple sources, we have reconstructed the key events surrounding this resignation.

On the evening of March 2, Lin Junyang led his team to release four small-sized open-source models of Qwen3.5. The models immediately attracted attention from the global technology community. Elon Musk commented on their "impressive intelligence density," and Lin Junyang himself retweeted the comment to express his gratitude. On the same day, Alibaba officially announced that it would unify the brand of its large-scale models for both B-end and C-end users under the name "Qianwen," establishing it as the group's core AI brand.

According to reports from LatePost and Phoenix Technology, on the afternoon of March 3, Lin Junyang left an internal meeting due to disagreements and submitted his resignation to Alibaba. After the news was shared, some colleagues from the Qwen team burst into tears.

On the evening of March 3, Lin Junyang shared the song "A Toast to Myself" on his WeChat Moments.

In the early hours of March 4th, Lin Junyang publicly announced his resignation on the X platform. As of 4 PM today, the post has received over 10,000 likes and more than 1,400 comments. Many industry developers left messages thanking him and the Qwen team for their contributions to the open-source field, and official accounts of AI companies such as MiniMax also appeared in the comments section to express their recognition.

On the morning of March 4, Yu Bowen, the head of Qwen post-training, and Li Kaixin, a core contributor to Qwen3.5/VL/Coder, announced their departure from the company.

On the afternoon of March 4th, Lin Junyang posted on his WeChat Moments: "Sorry everyone, I won't be replying to messages or phone calls today. I really need to rest. My brothers at Qwen, keep up the work as planned, no problem."

It is worth noting that Chen Cheng, a core contributor to the Qwen team, said frankly when forwarding Lin Junyang's post: "I am really heartbroken. I know leaving is not your choice. Just last night, we released the Qwen 3.5 miniature model together. I can't imagine what Qwen would be like without you."

This comment quickly fueled speculation within the industry that Lin Junyang's resignation was involuntary. As of now, Alibaba has not officially responded to this personnel change. An insider close to Alibaba believes that Lin Junyang's resignation was not voluntary, and his quick announcement of his departure on the X platform likely indicates that he couldn't control his emotions at the time.

Why did you leave the company?

Lin Junyang's sudden departure caused a huge uproar in the industry, with numerous speculations emerging. At its root, it was a concentrated outbreak of multiple long-term contradictions within Alibaba's organizational structure, technological roadmap, business goals, and talent pool.

The trigger may have been the restructuring of the team. According to LatePost, Tongyi Labs plans to break the vertically integrated R&D model of the Qwen team, splitting the original closed-loop team covering the entire process of pre-training, post-training, multimodal R&D, and infrastructure construction into multiple independent horizontal division modules, all directly under the coordination of Tongyi Labs. As a result, Lin Junyang's management authority and business boundaries will be narrowed.

Behind the organizational restructuring lies a fundamental conflict between Lin Junyang and Alibaba's senior management regarding their philosophies on large-scale model development. Lin Junyang believes that the core competitiveness of large-scale model development comes from deep collaboration across the entire process team. He argues that fragmented, pipeline-style divisions of laboratories severely deplete R&D efficiency and innovation space, and his team has already established its own self-developed infrastructure system. The proposed restructuring, which involves dismantling and reassembling components in the laboratory, completely contradicts his assessment of the development trend of large-scale model technology.

Even more difficult to reconcile than the differences in technical approaches is the deep-seated conflict between the open-source approach and the group's commercialization goals. Under Lin Junyang's leadership, Qwen topped the global open-source large model list with its comprehensive open-source strategy, becoming a benchmark for Chinese large models going global. However, Alibaba's core assessment of Qwen has shifted from building technological influence to commercialization.

However, internal doubts persist regarding the revenue efficiency of the open-source model. Reportedly, some executives even described Qwen-3.5, which debuted on New Year's Eve, as an unfinished "half-finished product." Coupled with the fact that the 3 billion yuan subsidy for the C-end Qianwen App fell short of expectations and the market pressure from aggressive competitors in the B-end AI cloud business, the misalignment between technological ideals and business goals became increasingly difficult to reconcile.

Behind this change, the restructuring of the talent landscape at Tongyi Lab has also led to a dilution of core influence. Since 2025, Alibaba has been continuously attracting top global AI talent. After IEEE Fellow Xu Zhuhong transferred to Tongyi Lab, the business he was responsible for largely overlapped with Qwen's layout. In early 2026, Zhou Hao, a former senior researcher at DeepMind, joined the company, reporting directly to the head of the lab, and will also take over the work of Yu Bowen, the former head of post-training.

The laboratory shifted from a single-core model led by Lin Junyang to a "multi-strong parallel" structure, coupled with the successive departures of Qwen's core founding team, which led to this industry-shaking departure event.

Who can replace Lin Junyang?

The reason why this departure has caused a chain reaction within Alibaba and the global open source community is due to Lin Junyang's key industry position in the Qwen team, Alibaba's AI system, and even in the field of global open source big models.

As a rare, entirely homegrown technology leader in China's AI industry, Lin Junyang, born in 1993, has an interdisciplinary background with a bachelor's degree in computer science and a master's degree in linguistics and applied linguistics from Peking University, which gives him a unique advantage in the development of large language models.

After graduating with a master's degree in 2019, he did not choose to go abroad to further his studies, but instead joined Alibaba DAMO Academy directly. Starting as a senior algorithm engineer, he rose through the ranks four times in six years and became Alibaba's youngest P10-level technical expert at the age of 32. He joined a group of technical leaders with overseas doctoral backgrounds, including Tang Jie of Zhipu AI, Yang Zhilin of Lunar Dark Side, and Yao Shunyu of Tencent, and was recognized in the industry as one of the "Four Great Masters of Basic Models" in the field of large-scale models in China.

When Lin Junyang took over as the technical lead for Tongyi Qianwen at the end of 2022, the large-scale model market in China was just emerging, and Alibaba's large-scale model strategy had not yet formed a clear differentiated advantage. Under his leadership, Qwen launched a comprehensive open-source strategy, creating a model family covering all parameter sizes from 0.8B to 72B. The flagship model Qwen3-Max with trillion parameters, launched in 2025, surpassed mainstream international models of the same period in several authoritative evaluations, including GPQA.

As of January 2026, the Qwen series of models has surpassed 200,000 derivative models on Hugging Face, the world's largest open-source AI community, with over 1 billion downloads, firmly holding the top position globally among open-source large models. Stanford University's "2025 AI Index Report" shows that the performance gap between top AI models from China and the US has narrowed to 0.3%, with Qwen's core model contribution ranking third globally.

After Lin Junyang turned and left, two questions arose from the outside world.

Who will succeed Lin Junyang? According to insiders, due to the suddenness of the incident, there is currently no candidate who can completely replace Lin Junyang. His original full-stack management work will be split up and distributed to multiple parallel teams along with the team.
Where will Lin Junyang go? An industry insider who has long observed Alibaba said that there is a high probability that he will start his own business, or he may join the team of embodied or world model stars. The probability of him being retained is very small.

For Alibaba AI, the primary impact of this personnel change is the chain reaction of core team departures and a decline in team morale. According to Phoenix Technology, Alibaba executives are still communicating with Lin Junyang to retain him.

Within just three months, Qwen's early core founding team, including its technical lead, post-training lead, and code lead, have all left the company. This will not only directly affect the subsequent iteration pace of Qwen's models, but may also lead to further talent loss.

Lin Junyang's departure marks a significant turning point in Alibaba's AI strategy, signifying a shift in priorities. This means that Alibaba's large-scale AI platform will move beyond its initial focus on building technological benchmarks and a global open-source ecosystem, and fully transition to a new cycle centered on commercialization.

However, Alibaba faces challenges such as disruptions to its R&D pace due to the loss of its core team, fluctuations in trust within the global open-source ecosystem, and fierce industry competition from rivals like ByteDance and Tencent. The chain reaction of this personnel shake-up will directly test the resilience of Alibaba's All in AI strategy.

Alibaba releases Fun-ASR 1.5, a large-scale speech recognition model.

Popular Articles