Baichuan AI Completes 5 Billion RMB Series A Financing, Valuation Exceeds 20 Billion RMB

According to a report by Chinese media outlet 36Kr, Baichuan AI has recently completed a Series A financing round, raising a total of 5 billion RMB.

Prior to this, Baichuan AI had announced its Series A1 financing round in October 2023, disclosing a list of investors that included tech giants such as Alibaba, Tencent, and Xiaomi, as well as several top-tier investment institutions. In the Series A financing round, in addition to the aforementioned institutions, Baichuan AI also received investments from state-backed industry investment funds, including the Beijing Artificial Intelligence Industry Investment Fund, the Shanghai Artificial Intelligence Industry Investment Fund, and Shenzhen Capital Group.

An insider revealed that after the completion of the Series A financing round, Baichuan AI’s valuation has reached 20 billion RMB, and the company will start its Series B financing round at this valuation.

Baichuan AI is one of the first large model companies established in China, founded in April 2023 by Wang Xiaochuan, former CEO of Sogou. Its core team consists of AI talents from tech companies such as Sogou, Google, Tencent, Baidu, Huawei, Microsoft, and ByteDance.

Wang Xiaochuan’s previous company, Sogou, was born during a period of intense competition in the search engine sector, and from day one, it had to seek a balance between reality and vision. For this reason, Wang Xiaochuan announced a dual-drive development strategy of “super model + super application” after establishing Baichuan – aiming to develop both AGI (Artificial General Intelligence) and specific scenarios.

In terms of both technology and application implementation, Baichuan’s pace can be described as quite fast. Since its establishment, Baichuan AI has released 12 large models, including Baichuan7B/13B, Baichuan2-192K, Baichuan-NPC, Baichuan 4, etc., covering both open-source and closed-source fields.

In May of this year, Baichuan also launched an AI assistant called “Bai Xiaoying,” equipped with multi-step search, directional search capabilities, and more. This is also a product in the intensely competitive To C AI products race in 2024.

However, for Baichuan, the AI assistant is just a transitional product. Compared to other large model players, Baichuan stands out for its strong belief in the medical field.

“Medicine is the jewel in the crown of large models,” Wang Xiaochuan has repeatedly stated in public. He believes that after breakthroughs are made in AI medical applications, the model size can be expanded based on real needs, the model’s capabilities can be improved, and then the upgraded model can be used to feed back into applications.

Medicine is also a direction that is receiving a lot of attention globally for the application of large models. Geoffrey Hinton, one of the three AI giants, mentioned in an interview in May this year that “medicine will be one of the most important fields of AI application.” He believes that if AI surpasses humans in some tasks, people could potentially have three doctors per person.

In terms of applying large models to medicine, Baichuan has disclosed quite a bit of progress this year. In January, Baichuan’s Baichuan 3 has already surpassed GPT-4 in several authoritative medical evaluation tasks.

At the recent World Artificial Intelligence Conference, Baichuan also released a beta version of an AI health consultant. Like a general practitioner with many years of experience, it can continuously converse with users, collect enough symptom information, and then make a comprehensive judgment, providing diagnostic results and medication suggestions.

Recently, Baichuan, benchmarking against autonomous driving, has given an L0-L5 classification for the technical development roadmap of AI in medicine. They believe that the current technology can achieve the L3 level – AI automatically recommends treatment plans under specific conditions, with key decisions confirmed by doctors. Because the demand in the medical field is deep enough, it can continue to AGI (Artificial General Intelligence) and even ASI (Super Artificial Intelligence) stage, eventually building a life-size model.

SEE ALSO: BYD Secures Order for 120 Electric Buses in South Africa