Tsinghua Expert Zhang Peng Leads Zhipu AI, Secures Over 2.5 Billion Yuan in Funding This Year
-
Investment circles learned that (October 20) Zhipu AI announced it has raised over 2.5 billion yuan in funding this year, with a star-studded lineup of investors—
The Social Security Fund Zhongguancun Independent Innovation Fund (with Legend Capital as the fund manager), Meituan, Ant Group, Alibaba, Tencent, Xiaomi, Kingsoft, Shunwei Capital, Boss Zhipin, TAL Education Group, Sequoia Capital, Hillhouse Capital, and other institutions, along with follow-on investments from existing shareholders including Legend Capital.
This is a large model company that emerged from a Tsinghua University laboratory, bringing together a group of Tsinghua elites—CEO Zhang Peng graduated from Tsinghua's Department of Computer Science, President Wang Shaolan holds a Ph.D. in Innovation Leadership from Tsinghua, and Professor Tang Jie from Tsinghua's Department of Computer Science also participated in its incubation.
Coincidentally, Wang Xiaochuan's large model company, Baichuan Intelligence, also announced this week that it has completed a $300 million funding round, another Tsinghua-affiliated venture. Looking around, this year has seen Tsinghua-affiliated founders in the AI field almost queuing up to announce funding. At the same time, AI entrepreneurs from Shanghai Jiao Tong University and the University of Science and Technology of China are equally active, collectively creating a lively scene in the AI world.
The story of Zhipu AI began in 2006 at Tsinghua University's laboratory.
That year, a team from Tsinghua's Knowledge Engineering Laboratory released a product called AMiner for mining scientific research intelligence. This technology quickly gained a group of initial clients upon release, with the project being led by Tang Jie, a renowned professor from Tsinghua's Computer Science Department.
In 2019, CEO Zhang Peng and his team emerged from Tsinghua Laboratory, and Zhipu AI was officially established.
Zhipu AI CEO Zhang Peng
From its inception, Zhipu AI attracted investments from prominent institutions such as CAS Investment, Tsinghua Holdings, and Huakong Huijin. With a technically proficient team, co-founder Zhang Peng stated the company launched with ready technology, team, and market access: "We developed core technology while expanding our market. We generated revenue from day one."
In 2020, OpenAI's GPT-3 release overseas began disrupting the AI industry. The emergence of GPT convinced Tang Jie and Zhang Peng that large models represented the future direction. On Zhipu AI's first anniversary, the team committed fully to large model development. By August 2022, GLM-130B was officially launched.
Soon after, Zhipu AI gathered a group of renowned venture capital firms behind it. Just one month after the release of GLM-130B, Zhipu AI secured a 100 million yuan Series B funding round co-led by Legend Capital and Qiming Venture Partners. Prior to this, the company had already completed a funding round exceeding 100 million yuan with participation from Fortune Capital, HuaKong Fund, Jiangmen Venture Capital, Nanjing Turing, and Beijing Dafan, among others.
With the explosive popularity of ChatGPT, large language models began entering the public consciousness. In July of this year, Meituan's strategic investment arm led a B-2 funding round for Zhipu AI exceeding $100 million, valuing the company at $500 million (approximately 3.6 billion yuan), marking the beginning of Zhipu AI's massive funding spree this year. Over the following months, Zhipu AI successively secured multiple funding rounds amounting to hundreds of millions or even billions of yuan.
Ultimately, Zhipu AI has secured over 2.5 billion RMB in funding this year, with participation from major institutions including the Social Security Fund Zhongguancun Independent Innovation Fund (managed by Legend Capital), Meituan, Ant Group, Alibaba, Tencent, Xiaomi, Kingsoft, Shunwei Capital, Boss Zhipin, TAL Education Group, Sequoia Capital, Hillhouse Capital, and follow-on investments from some existing shareholders, including Legend Capital.
Despite being a startup, Zhipu AI has also been actively investing. In the field of large models, Zhipu AI has participated in the angel and Pre-A funding rounds of Lingxin Intelligence. Additionally, it has invested in another Tsinghua-affiliated large model company, Mianbi Intelligence, whose co-founder is Liu Zhiyuan, the deputy director of the Tsinghua Foundation Model Research Center.
Looking back at the past four years, large models have undergone earth-shaking changes. At its inception, the Zhipu AI team wrote down a vision—'to make machines think like humans.' Today, this vision is becoming a reality.
This week marked the second time Alibaba, Tencent, and Xiaomi have joined forces in a rare collaboration.
On October 17, Baichuan Intelligence announced the completion of a $300 million Series A1 strategic financing round, with participation from tech giants such as Alibaba, Tencent, and Xiaomi. The company has officially joined the ranks of tech unicorns, setting a record for the fastest promotion to unicorn status among domestic large model startups. Additionally, institutions like Shenzhen Capital Group, Cornerstone Capital, Redpoint China, Zhuoyuan Capital, and Shunwei Capital also participated in this round.
The founder of Baichuan Intelligence is no stranger to the venture capital circle—Wang Xiaochuan. In 1996, Wang Xiaochuan was admitted to the Computer Science Department of Tsinghua University and joined Sohu after completing his postgraduate studies. From Sohu to Sogou, Wang Xiaochuan left many marks in the internet era. It wasn't until April of this year that he officially announced his venture into large model development, aiming to create a Chinese version of OpenAI.
Before Wang Xiaochuan, fellow Tsinghua alumnus Wang Huiwen was the first to disrupt the calm in the large model industry. Earlier this year, Wang Huiwen announced his entry into the artificial intelligence field, recruiting talent for Lightyear Beyond, an AI company he co-founded, with an initial personal investment of $50 million. As his Tsinghua classmate, Wang Xing strongly supported the venture, participating in Lightyear Beyond's Series A funding and joining as a board member.
During this period, Lightyear Beyond also merged with OneFlow Technology, which was founded in 2017. OneFlow's founder and CEO Yuan Jinhui is also a Tsinghua alumnus, holding a Ph.D. in Engineering from the university's Computer Science Department and previously serving as a lead researcher at Microsoft Research Asia. Unfortunately, Wang Huiwen was reported to have left his position due to illness at the end of June, and Lightyear Beyond was subsequently fully acquired by Meituan.
Also drawing significant attention is the large model startup MoonDark - led by Professor Yang Zhilin, a young scientist from Tsinghua University's Institute for Interdisciplinary Information Sciences and Zhiyuan, with two co-founders Zhou Xinyu and Wu Yuxin also Tsinghua alumni. The investment community has learned that the company has received nearly 2 billion yuan in funding from renowned institutions including Sequoia Capital, Today Capital, Linear Capital, and ZhenFund.
Why are Tsinghua-affiliated entrepreneurs so dominant in the AI field?
This story dates back 45 years. As early as 1978, Tsinghua University established the "Artificial Intelligence and Intelligent Control" teaching and research group within its Computer Science Department, offering undergraduate AI education under the "Computer Application Technology" discipline. In 1979, Tsinghua began offering the elective course "Introduction to Artificial Intelligence" for undergraduates. By 1983, the Computer Science Department revised its undergraduate education plan, making "Artificial Intelligence" a compulsory course.
Starting in 1978, Tsinghua began enrolling master's students in artificial intelligence, with the first cohort consisting of six students: Zhang Guoxuan, Zhang Zhoucai, Wang Jiaqin, Zheng Xuezhong, Tang Jianbang, and Yan Juanwei, all of whom later became key figures in their respective fields. In 1986, Tsinghua also began enrolling doctoral students in AI, and by 2020, it had cultivated approximately 250 PhD graduates in this field.
Tsinghua University once organized three phases of its Distinguished Professorship Program, inviting world-renowned professors to participate in the university's talent cultivation efforts. The main activities included teaching and joint supervision of doctoral students, with the requirement that each professor group spend a cumulative total of nine months per year working at Tsinghua University. The first phase ran from 2003 to 2006 and primarily involved professors in theoretical computer science, including Turing Award winner Andrew Yao. Later, Yao accepted a permanent professorship at Tsinghua University and established the Yao Class in 2005, which has been hailed as the cradle of China's artificial intelligence industry.
According to Wang Xingshi, a partner at Source Code Capital and also a Tsinghua alumnus, AI entrepreneurs from Tsinghua possess a strong vision for technological innovation, excel at deeply exploring the essence and patterns of things, and maintain a pragmatic approach in practice.
This is inseparable from Tsinghua's underlying culture. Wang Xingshi said that Tsinghua has always advocated pragmatism and innovation, adhering to the concept of continuously cultivating outstanding high-tech talents for the country and encouraging entrepreneurship and innovation. Therefore, under the influence of this overall culture and atmosphere, it is easier to foster and stimulate entrepreneurial awareness. "When opportunities for entrepreneurship and innovation arise in industries like AI, students with strong professional knowledge, innovative thinking, and entrepreneurial spirit will seize these opportunities, leading to the emergence of many top entrepreneurs."
As the saying goes, "Tsinghua in the north, Jiao Tong in the south," the Shanghai Jiao Tong University community is also a formidable force in the AI arena.
The latest development occurred on September 28th when 4Paradigm officially listed on the Hong Kong Stock Exchange, achieving a market capitalization exceeding HKD 26 billion. The founder of 4Paradigm is Dai Wenyuan, a young academic prodigy from Shanghai Jiao Tong University. Passionate about computers from an early age, Dai secured direct admission to Shanghai Jiao Tong University through competitions and studied in the prestigious ACM class. During his university years, he won the ACM World Championship and three Asian championships. After graduation, Dai became Baidu's youngest scientist at the time and went on to establish 4Paradigm in September 2014.
Less known is the fact that 4Paradigm's Executive Director and Chief Research Scientist Chen Yuqiang, Chief Architect Hu Shiwei, and Vice President Zheng Zhao are also alumni of Shanghai Jiao Tong University.
There is also Xu Li, co-founder and current CEO of SenseTime. In 2000, he was admitted to the Computer Science Department of Shanghai Jiao Tong University through a recommendation program, and four years later, he was directly recommended for graduate studies. Subsequently, he earned his Ph.D. in Computer Science and Engineering from The Chinese University of Hong Kong. At CUHK, Xu Li met Professor Xiaoou Tang and co-founded SenseTime in the laboratory.
Looking back at history, Shanghai Jiao Tong University was one of the earliest universities in China to venture into the field of AI. As early as 1973, Shanghai Jiao Tong University established one of the earliest and largest computing centers in Chinese universities—the Shanghai Jiao Tong University Network Information Center—and gradually built a top-tier AI computing platform in the country.
The most prestigious program at Shanghai Jiao Tong University is undoubtedly the ACM class, which was established in 2002 by Professor Yu Yong. During ByteDance's 8th anniversary, Zhang Yiming mentioned the ACM class in an internal letter: "At one point, I noticed that many outstanding algorithm talents in our company came from Shanghai Jiao Tong University's ACM class, so I made a special trip to Shanghai to visit Professor Yu Yong."
This class has produced a group of legendary figures in the computer science field. According to incomplete statistics, graduates from the ACM class are spread across various artificial intelligence domains, including Lin Chenxi, co-founder of Yitu Technology; Xu Lingjie, co-founder of Biren Technology; and Sun Kai, chairman and chief scientist of Hesai Technology. Without a doubt, artificial intelligence has become a flagship program at Shanghai Jiao Tong University.
In the AI arena, there is another university that shines brightly—the University of Science and Technology of China (USTC), located in Hefei.
The most representative case is Cambricon Technologies. The company's founders, brothers Chen Yunji and Chen Tianshi, both studied in USTC's Young Scholars Program and are renowned young scientists in academia both domestically and internationally. In 2016, the brothers co-founded Cambricon Technologies, which became the first AI chip unicorn company and successfully went public on the STAR Market in 2020.
Cloudwalk Technology, one of the 'AI Four Tigers', is also an enterprise founded by alumni of the University of Science and Technology of China (USTC). Zhou Xi, born in 1981 in Neijiang, Sichuan, successively enrolled at USTC and the University of Illinois at Urbana-Champaign in the United States, where he earned a Ph.D. in Computer Science. In 2011, Zhou Xi chose to return to Chongqing to research AI instead of staying in the U.S., and Cloudwalk Technology was born four years later. In May of last year, Cloudwalk Technology was officially listed on the STAR Market.
USTC alumni active in the AI field are too numerous to count: Academician Zhang Yaqin of the Chinese Academy of Engineering, Tang Xiao'ou founder of SenseTime, Liu Qingfeng founder of iFlytek, former Alibaba Cloud Chief Scientist Min Wanli, Liang Jia'en founder of Unisound... The list is impressive.
The road ahead is long and arduous, but I will search high and low. Undoubtedly, China needs to give birth to its own OpenAI, and will eventually have its own OpenAI. On this path of knowledge, the Chinese AI corps from major universities are gathering in full force, marching together toward the distant future.