According to two persons familiar with the situation, Beijing-based AI firm 01.AI is collecting up to $200 million in new cash, expanding on a $1 billion valuation attained last month as a result of rising worldwide interest in open-source AI models.

It is one of many Chinese businesses that have released or intend to expose their large language models (LLM) to the public, joining bigger corporations Meta and Alibaba in a race to recruit users and catch up to industry leader OpenAI.

According to one of the sources, 01.AI, which was formally established in July by Lee Kai-fu, Google China’s former CEO, following a three-month incubation phase, reached a $1 billion value early in November.

According to another source, the corporation is looking for more investments in US dollars.

Because the funding information was not made public, all of the sources quoted in this piece requested anonymity.

A request for comment was not immediately returned by the corporation.

In November, 01.AI gained attention in the open-source LLM community when their Yi-34B model became the first Chinese LLM to top the scoreboard on Huggingface, a platform for IT businesses to exchange LLMs, which are then graded for performance and popularity.

Riding the wave of interest in generative artificial intelligence (AI), such businesses have raised large funding from investors after pledging to open up their AI models.

According to Chinese media, the value of Zhipu AI, which was created in 2019, has surpassed $1 billion.

According to a third source with firsthand knowledge of the transaction, Baichuan Technology was valued at $1.2 billion following its most recent financing in October.

SKEPTICS FROM OPEN SOURCES

The open-source LLM movement has been chastised by OpenAI, the creator of the popular chatbot ChatGPT, which has kept the codes of its models under wraps, raising the potential of exploitation by bad actors, which might endanger society.

Alibaba’s cloud subsidiary in China has been aggressive in making their LLMs open-source. Its most current LLM Qwen-72B model just topped Huggingface’s leaderboard, becoming the second Chinese model to do so.

Qwen-72B is the most recent and powerful of eight AI models released by Alibaba in the previous four months. According to certain industry benchmarks, the Qwen-72B can exceed OpenAI’s flagship GPT4 in handling Chinese. 

“Creating an open-source ecosystem is critical to promoting the development of LLM and AI applications,” said Jingren Zhou, Alibaba Cloud’s chief technology officer, to Reuters.

He noted that demand for AI applications is expanding across a broad spectrum of sectors, developers, and enterprises.

“Alibaba Cloud aspires to become the most open cloud and make generative AI capabilities accessible to everyone.“

The business also helps smaller open-source AI LLM companies, like as 01.AI, with whom it collaborates on model training and deployment procedures.

According to the sources, AliCloud has already committed funds in the current round.

AliCloud did not reply quickly to a request for comment.