Large language model technology is rapidly developing globally, and Chinese companies are actively following this trend, applying large model technology in multiple fields to improve operational efficiency and user experience. However, large model enterprises still face many challenges, including technological maturity, costs, data security, industry differences, and user acceptance. Commercialization remains a key issue for all vendors to solve.
Although the commercial prospects of large models in the To C field are still unclear, they have begun to show scale in the bidding market. Large model technology is gradually penetrating more industries, with related projects appearing in fields such as rice technology, weather forecasting, and medical pre-diagnosis.
According to data, in 2023, the bidding market initiated 190 large model procurement demands, with a procurement scale of 595 million yuan. In the first half of 2024, related bidding projects have reached 498, with a bidding amount exceeding 1.34 billion yuan, already surpassing the full-year scale of 2023.
In terms of the number of successful bids by suppliers, ZhipuAI, Baidu, iFlytek, and Huawei are at the forefront. Here are the bidding situations of some major suppliers:
-
ZhipuAI: Won 18 projects, covering finance, education, energy, communications, healthcare, and other fields.
-
Baidu: Won 17 projects, involving healthcare, finance, energy, environmental protection, and transportation.
-
iFlytek: Won 15 projects, covering finance, education, energy, communications, justice, and other fields.
-
Huawei: Won projects covering energy, finance, meteorology, and other industries.
-
SenseTime: Winning projects mainly concentrated in important fields such as electricity and finance.
-
Alibaba Cloud, Tencent Cloud, and Volcano Engine: As infrastructure providers, they have relatively fewer publicly won projects.
It's worth noting that some large model startups such as MiniMax, Moonshot AI (and its affiliated company Recurrent AI), ZeroOne.AI, and Baichuan Intelligence appear less frequently or not at all in public bidding information. This may be related to these companies' development strategies or market positioning.
Overall, large model technology in China is gradually moving from concept to practical application but is still in its early stages of development. Different companies have adopted different commercialization strategies based on their strengths and market positioning. In the future, as the technology further matures and application scenarios expand, large model technology is expected to play an important role in more fields.