Chinese tech giants slash LLM prices

Competition in China's artificial intelligence-powered large language model sector is intensifying as leading Chinese tech companies have announced drastic price reductions on major LLM products, leading to a growing price war, industry experts said.

They added the application of generative AI is expected to see speedy growth along with technological advances in LLMs. They highlighted that the new round of price cuts will accelerate the commercial use and popularization of AI technology, as well as bolster and upgrade industrial development.

Alibaba Cloud, the cloud computing unit of Alibaba Group Holding Ltd, announced on Tuesday a price reduction of 97 percent on its Qwen-Long LLM, which is equivalent to the performance of US company OpenAI's GPT-4 model.

The model, which supports up to 10 million tokens of long text input, now only costs about one-400th of the GPT-4 price. A token is the smallest unit into which text data can be broken down for an AI model to process.

The steep price cuts cover nine of Alibaba's self-developed LLMs known as Tongyi Qianwen. Liu Weiguang, president of public cloud business at Alibaba Cloud Intelligence, said the purpose of slashing prices this time is to speed up the application of AI models in the domestic market.

Hours after Alibaba Cloud's announcement, Chinese tech heavyweight Baidu Inc said its LLMs — including Ernie Speed and Ernie Lite, both launched earlier this year — are now available free of charge for all enterprise users. Last week, ByteDance announced that the main model of its Doubao LLMs will be priced 99.3 percent lower than the industry average for business users.

Chinese AI firm iFlytek said on Wednesday some versions of its SparkDesk LLM are free, or five times cheaper than similar products from competitors. Tencent Holdings also announced price cuts on its Hunyuan LLMs, and the Hunyuan-lite model is available for free from Wednesday.

Chinese LLM companies have already made great progress in optimizing algorithms and improving computing power, and are able to provide AI-powered LLM services at a lower cost, said Wang Peng, a researcher at the Beijing Academy of Social Sciences.

Wang said the substantial price reductions for LLMs will not only lower the threshold for enterprises to adopt AI technology, but also promote their digital transformation, adding that the application of AI technology will help enterprises improve operational efficiency and reduce costs, and further enrich the ecosystem of AI applications.

Experts said Chinese LLM developers have primarily targeted businesses for monetizing their investments in AI models, although some have begun targeting individual users who need to pay for using advanced chatbot services.

Alibaba said earlier this month that its LLMs have been adopted by more than 90,000 corporate clients in a wide range of industries, one month after Baidu said its enterprise-level LLM platform Qianfan had been used by 85,000 business users.

Foreign AI model developers are also reducing their LLM prices. OpenAI has slashed prices on its products four times since the beginning of 2023, and the recently unveiled GPT-4o is 50 percent cheaper than the GPT-4 Turbo.

"The technological strengths, application scenarios and training costs of AI models vary among different LLM companies, which leads to differences in pricing tactics," said Zou Debao, deputy general manager of AI and big data research at market research firm CCID Consulting.

Zou said the price war will likely squeeze profit margins of some LLM startups, while large Chinese LLM developers will further consolidate their dominant positions, expand market share and grab more clients amid increasingly fierce competition.

Official data showed that the number of China's self-developed LLMs ranks second in the world, and over 100 LLMs with more than 1 billion parameters each have been launched in the country.

Zhao Gang, head of CIO Manage, a Beijing-based digital economy market consultancy, said the continuous increase in LLM users and expansion of application scenarios will be conducive to improving the performance of LLMs.

The significant AI model price cuts will give a boost to the rapid growth of the LLM sector, and motivate companies to beef up innovation efforts to roll out more LLMs for commercial applications, Zhao said.