Former Lead of Qianwen Seeks Hundreds of Millions in Funding for New AI Lab

According to reports from The Information citing two informed sources, the former lead of Alibaba's Qianwen Model, Lin Junyang, is seeking hundreds of millions of dollars in financing for a new AI lab, with a post-financing valuation likely to be around $2 billion. GaoRong Ventures and Sequoia China are in talks to participate in the investment, but the deal has not been completed, and the final valuation may still change.

This valuation is not common among Chinese AI startups. The name of the new lab has not been disclosed yet, and there are limited externally verifiable products and commercialization progress. At this stage, investors are mainly betting on Lin Junyang's technical background at Alibaba, as well as his experience in leading Qianwen to the forefront of global open-source AI.

Lin Junyang publicly announced his departure from Alibaba at the beginning of March this year, shortly after the release of Qwen 3.5. Subsequently, Alibaba integrated the Qwen research team and other key AI businesses into a new department directly managed by the company's CEO, and shifted towards providing more proprietary models that can generate revenue.

This round of financing also occurs against the backdrop of a resurgence in Chinese AI assets. After listing in Hong Kong earlier this year, ZhiPu and MiniMax saw their stock prices rise. DeepSeek is reported to be seeking external financing for the first time and has raised its expected valuation multiple times during negotiations.

However, high valuations do not guarantee smooth transactions. Venture capitalists believe that the high valuations of American research-oriented AI startups are partly due to the expectation of potential acquisitions by large technology companies in the future, and it remains uncertain whether this exit logic can be replicated in China. Additionally, the export controls imposed by the United States on advanced AI chips to China necessitate that the new lab address computational resource issues, while also finding a research direction that does not completely overlap with those of major factories like Alibaba and ByteDance.