On Tuesday, Chinese tech giants Alibaba and Baidu announced significant price reductions for their large language models (LLMs) that power generative artificial intelligence products such as chatbots, intensifying the price war in China’s cloud computing sector.
Alibaba’s cloud division revealed price cuts of up to 97% on various Tongyi Qwen LLMs. For example, the cost of their Qwen-Long model has been slashed to just 0.0005 yuan per 1,000 tokens, a significant drop from the previous price of 0.02 yuan per 1,000 tokens.
Following Alibaba’s announcement, Baidu quickly responded by making its Ernie Speed and Ernie Lite models free for all business users.
This move is part of a broader price competition in China’s cloud computing market, which has seen Alibaba and Tencent recently reducing prices for their cloud services. Chinese cloud vendors have increasingly turned to AI chatbot services to drive sales, following the substantial interest in large language models sparked by the success of OpenAI’s ChatGPT in late 2022.
The current price war now encompasses the large language models that support these chatbots, potentially squeezing profit margins for companies involved. Baidu’s Ernie Lite and Ernie Speed, introduced in March, were previously accessible to corporate customers on a paid basis until the recent price adjustment. Last week, Bytedance also announced that the primary model of its Doubao LLMs would be priced 99.3% lower than the industry average for business users.
Chinese developers of LLMs have primarily focused on monetizing their investments by charging businesses. However, some are now targeting individual users as well. For instance, Chinese startup Moonshot recently introduced a tipping feature, allowing both business and individual users to pay for prioritized access to its chatbot services. Baidu pioneered offering its LLM products to paying consumers in China, charging 59 yuan per month for access to its advanced Ernie 4 model.
Written by Alius Noreika