Chinese tech giants continue to slash prices of LLMs to power AI chatbots

Alibaba Cloud, the cloud computing division of the Chinese e-commerce firm, announced a significant price reduction of 85 percent on its visual language model Qwen-VL on Tuesday, a latest move showing that Chinese tech companies are in an intensifying race of business for their nascent artificial intelligence (AI) products.

This is the third round of price cuts following the two price cuts in May and September 2024 to incentivize businesses to use its AI products. 

In May 2024, Alibaba Cloud reduced the price of its Qwen-Long, which was its first significant price reduction that year. The API input price of the model dropped from 0.02 yuan ($0.0027)/thousand tokens to 0.0005 yuan/thousand tokens, a drop of 97 percent.

This time, Qwen-VL-Plus dropped by 81 percent, with an input price of only 0.0015 yuan per thousand tokens, the lowest price online, and the higher-performance Qwen-VL-Max dropped to 0.003 yuan per thousand tokens, a drop of up to 85 percent. According to the latest price, 1 yuan can process up to about 600 720P pictures, or 1,700 480P pictures, according to the news outlet chnfund.com on Tuesday.

This price reduction is mainly due to the continuous optimization of Alibaba Cloud's infrastructure and model structure, as well as the scale effect brought about by the exponential growth of large language models (LLMs) calls, according to the information Alibaba Cloud shared with the Global Times on Wednesday. 

Major Chinese tech firms including Alibaba, Tencent, Baidu, Huawei and TikTok parent company Bytedance have all joined the price war of their own large language models over the past year, looking to capitalize on the hype around the technology.

ByteDance has intensified competition in China's generative AI market by significantly reducing the cost of its new AI model.

The new model, part of the company's popular Doubao family, was introduced at 0.003 yuan per thousand token uses, reported the scmp.com on December 20, citing Tan Dai, president of ByteDance's Volcano Engine cloud unit, at a corporate event.

The reasons behind the price cuts involve changes in demand, market competition and policy support, Liu Gang, director of the Nankai Institute of Economics and chief economist at the Chinese Institute of New Generation AI Development Strategies, told the Global Times on Wednesday.

The reduction in costs is a critical factor in propelling AI models into the value-creation phase, indicating a rapid transition toward large-scale commercial use, according to Liu.

As AI becomes more affordable, its impact becomes more evident across various industries. "Lower costs make advanced AI tools accessible to businesses of all sizes, driving innovation and competition on a global scale," Liu said, adding that the price cut trend indicates a continued focus on affordability and accessibility, shaping the mainstream adoption of AI technologies.

The decrease in pricing for LLMs is advantageous for both application developers and industry players, which will enable the creation of a diverse range of innovative AI applications, Tian Feng, dean of SenseTime Intelligent Industry Research Institute, told the Global Times on Wednesday.