A post-1980s Tsinghua University professor secured a single-round financing of 2 billion yuan.

The largest single - round financing for a domestic video large - model has emerged.

Just today, Shenshu Technology, a general world - model AI enterprise, officially announced the completion of a nearly 2 - billion - yuan Series B financing, led by Alibaba Cloud.

It's worth mentioning that just two months ago, Shenshu Technology had just completed a Series A + financing of over 600 million yuan, surpassing the previous single - round financing record in the domestic video generation field held by Aishi Technology. Now, Shenshu Technology has refreshed its own record once again.

With the successful acquisition of 2.6 billion yuan in financing over the past two months, Shenshu Technology's valuation has risen from 8 billion yuan to 12 billion yuan, quickly entering the unicorn club.

Interestingly, this is not the first time that the Alibaba ecosystem has led an investment in Shenshu Technology.

As early as June 2023, just three months after Shenshu Technology was registered, Ant Group made its "first investment" in a large - model company with a nearly 100 - million - yuan angel - round investment in Shenshu Technology, betting on the long - term development potential of the video generation track.

With Alibaba Cloud leading the investment this time, combined with the fact that the mysterious AI video model "Happy Horse" dominated the charts this week and was just claimed by Alibaba, Alibaba is accelerating its entry into AI video generation through increased investment and self - research, and may become a new "catfish" stirring up the market.

An 80s - born Tsinghua professor leads the way to create a unicorn

Being heavily invested in by Alibaba twice, Shenshu Technology naturally has an impressive background.

Zhu Jun, the company's technical leader, was born in 1983. He is an IEEE Fellow, a current professor at the Department of Computer Science at Tsinghua University, the deputy director of the Institute for Artificial Intelligence at Tsinghua University, the director of the State Key Laboratory of Intelligent Technology and Systems at Tsinghua University, and the chief scientist of the Beijing Academy of Artificial Intelligence.

During his studies, he obtained both his bachelor's and doctoral degrees from the Department of Computer Science at Tsinghua University, under the tutelage of Zhang Bo, an academician of the Chinese Academy of Sciences and a professor at the Department of Computer Science at Tsinghua University.

Although Shenshu Technology was registered in 2023, its technological accumulation can be traced back to 2013.

Its core founding team is the research group led by Zhu Jun, one of the earliest international teams to study deep probabilistic generative models. They have long been deeply involved in the theoretical and algorithmic research of Bayesian machine learning and have accumulated profound technological expertise in areas such as diffusion models and multimodal fusion.

As early as 2021, when the industry's focus was still on large image - text models, Zhu Jun predicted that "the next breakthrough would be in video generation."

This prediction has become the core logic for Shenshu Technology to focus on the video generation track.

In September of the following year, the team pioneered the U - ViT architecture. As the world's first architecture to introduce Transformer into the diffusion model system, it can not only be used to generate high - quality images and videos but also 3D content and even to build world models, laying a solid foundation for Shenshu Technology's core technology.

In 2024, based on U - ViT technology, Shenshu Technology launched the large video - generation model Vidu, becoming China's first large - scale video - generation model product fully comparable to Sora, nearly a year earlier than ByteDance's Seedance.

It is reported that as the core product of Shenshu Technology, the Vidu series of text - to - video large models have completed three major technological iterations, gradually achieving a leap from "usable" to "easy - to - use" and then to "professional."

On January 30 this year, Shenshu Technology also released the Vidu Q3 model, focusing on professional - level film and television production scenarios. In the latest list published by the international authoritative AI benchmark testing institution Artificial Analysis, it ranked first in China and second globally, only behind xAI's Grok video - generation model.

In addition, according to the latest evaluation results of the authoritative video - generation model evaluation benchmarks VBench - 1.0 and VBench - 2.0, Vidu Q1 outperformed top domestic and international models such as Runway, OpenAI Sora, and Kuaishou's Kling on both VBench series lists, ranking first on both text - to - video track lists.

While Shenshu Technology is making rapid technological progress, it is also doing well in the primary market in terms of financing.

It completed three rounds of angel - round financing within a year, attracting well - known enterprises and VCs such as Ant Group, BV Baidu Ventures, Qiming Venture Partners, and Zhipu AI.

Subsequent Series A financing extended from the Pre - A round in June 2024 to the A + round in February this year, with more than six institutions participating in each round.

To date, Shenshu Technology has completed seven rounds of financing in three years, with its valuation soaring from 700 million yuan in the angel round to 12 billion yuan in the Series B round, a 16.14 - fold increase.

Notably, this year, it raised 2.6 billion yuan in just two rounds, and its valuation increased by 4 billion yuan in just two months.

Behind this new unicorn, Baidu participated in two rounds of investment, and Baidu Ventures participated in four rounds, showing the longest - standing support.

The Alibaba ecosystem, leading two rounds of investment, has firmly established itself as the most generous investor in Shenshu Technology.

The Alibaba ecosystem invests and conducts research simultaneously, stirring up the AI video - generation market

The investment from the Alibaba ecosystem not only injects "rocket fuel" into Shenshu Technology's development but is also an inevitable part of Alibaba's AI ecosystem strategy.

In fact, Alibaba is strategically accelerating its investment in AI video generation.

When making the angel - round investment in Shenshu Technology, Ant Group stated that video generation is a core track in the AIGC field and a key direction for the upgrade of the digital content industry. Shenshu Technology can synergize with Ant Group's digital ecosystem, helping Ant Group in its layout in areas such as AI content generation and digital marketing.

Subsequently, Alibaba made two consecutive investments in AI video generation. First, in September last year, Alibaba led a Series B financing of over 60 million yuan for Aishi Technology. Then, half a year later, Alibaba Cloud led a nearly 2 - billion - yuan Series B financing for Shenshu Technology.

The logic behind Alibaba Cloud's role as the main investor this time is quite simple.

Shenshu Technology has gradually built a complete application matrix including Vidu MaaS (Model as a Service), Vidu SaaS (Software as a Service), Vidu Agent, and Vidu APP, covering all scenarios such as B - end enterprise customers, C - end individual users, and professional - level creators, achieving full - scale penetration of technological commercialization.

According to public information, in 2025, the company achieved over a tenfold increase in both users and revenue, covering more than 200 countries and regions.

By the end of 2025, Shenshu Technology had reached cooperation agreements with many industry - leading customers such as JD.com, Amazon, Feishu, CCTV Animation, Anta, Honor, and Lenovo, with the highest industry - leading customer coverage rate in the industry.

This means that through investment, Shenshu Technology is not only the vanguard that enables Alibaba to enter the first echelon in the video - generation field after Aishi Technology but will also be the AI customer with the highest demand for computing power for Alibaba Cloud.

Additionally, according to media reports, Alibaba Cloud's relevant sales teams have started to expand their customer base in the fields of AI short dramas and comic dramas.

In addition to investment, Alibaba is also accelerating its self - research on AI video - generation models.

This week, an unnamed model called HappyHorse - 1.0, as a dark horse, topped the Artificial Analysis video arena ranking with 1333 Elo points and has been claimed by Alibaba.

Previously, there were reports that HappyHorse was developed by a team led by Zhang Di from the Future Life Laboratory of Taotian Group under Alibaba. Currently, the Future Life Laboratory has been spun off from Taotian Group and is now under the AI Innovation Division of the newly established ATH Business Group.

On March 16, the ATH (Alibaba Token Hub) Business Group was announced by Alibaba CEO Wu Yongming and is led by him personally. It integrates five major sectors: Tongyi Laboratory, MaaS business line, Qianwen Division, Wukong Division, and AI Innovation Division, concentrating Alibaba's core AI resources with the core goal of "creating tokens, delivering tokens, and applying tokens."

This means that Alibaba is aggregating its superior forces and resources and entering the most critical battlefield, marking the full - scale acceleration period of its AI development.

According to Goldman Sachs' prediction, the global market size of the AI video - generation industry will increase from approximately $3 billion in 2025 to approximately $29 billion in 2030, nearly a tenfold increase in five years.

As the industry enters the critical period of commercialization, Alibaba naturally doesn't want to miss out on the opportunity.

This article is from the WeChat official account "Lieyun Selection." Author: Sun Yuan. Republished by 36Kr with permission.