China's OpenAI challenger, Zhipu AI, has secured funding from the renowned food delivery giant Meituan, which currently has a market cap of approximately $100 billion. As reported in business filings, Meituan has become a 10% shareholder of Zhipu AI through its subsidiary.
According to TechCrunch, the exact funding amount remains undisclosed, but the startup previously raised "hundreds of million yuan" during a Series B round in September from investors like Qiming Venture Partners, Legend Capital, and Tsinghua Holdings.
China's OpenAI challenger, Zhipu AI, has secured funding from the renowned food delivery giant Meituan.Lintao Zhang/Getty Images
All About Zhipu AI of China
Zhipu AI, established in 2019, originated from the Knowledge Engineering Laboratory of Tsinghua University's Computer Science Department. Leveraging over a decade of technical expertise in knowledge intelligence, Zhipu AI has assembled a team of technology researchers and senior consultants.
Among them are CEO Zhang Peng, a Tsinghua University Computer Science graduate; Chairman Liu Debing, an apprentice of Academician Gao Wen from the Chinese Academy of Engineering; and President Wang Shaolan, who holds a doctorate as a Tsinghua Innovation Leader.
Notably, Zhipu AI has actively participated in the development of a large-scale pre-training model, integrating a high-precision general knowledge map, resulting in a cognitive engine driven by data and knowledge.
This core technology has been recognized with the second prize of the National Science Progress Award and the first prize of the Beijing Municipal Invention Patent Award.
Within three years, Zhipu AI introduced the concept of Model as a Service (MaaS) to the market, establishing a significant cognitive model platform alongside digital human and scientific and technological intelligence products.
Zhipu AI serves a vast clientele comprising over 1,000 enterprises and institutions, including prominent names like China Association for Science and Technology, Beijing Municipal Science and Technology Commission, Huawei, and Tencent.
Recently, Zhipu AI collaborated with academia and industry to create GLM-130B, an open-source high-precision bilingual 100 billion model supporting both Chinese and English languages.
The company claimed that this model surpassed GPT-3's performance on multiple public evaluation sets and efficiently operated on various chips like Nvidia, Huawei Ascend, Haiguang DCU, and Sunway supercomputer.
The company's collaboration with Tsinghua University and Huawei resulted in the release of CodeGeeX, a code generation model enabling automatic code generation with minimal comments and seamless conversion between different programming languages.
With Meituan's backing and continued funding, Zhipu AI aims to advance its technologies and potentially present a challenge to OpenAI's current dominance in the AI field.
