Highlights:

  • The price reductions show how fiercely China’s largest tech companies are competing to get more customers for their fledgling artificial intelligence products.
  • Alibaba Cloud reported in May that its Qwen models had been downloaded by over 90,000 businesses nationwide.

Alibaba Cloud slashed the cost of access to its most advanced large language models (LLMs) by up to 85% to grab more attention from Chinese enterprises.

Reports claim that the Qwen-VL-Max model from Alibaba Cloud is currently only 0.003 yuan (USD 0.00041) per thousand tokens. With its increased pricing, the visual reasoning model Qwen-VL-Max, which can sense and comprehend both textual and image-based inputs, significantly outperforms competing models from firms like ByteDance Ltd.

The cloud computing division of the massive Chinese e-commerce company Alibaba Group Holding Ltd. is called Alibaba Cloud. The price reductions show how fiercely China’s largest tech companies are competing to get more customers for their fledgling artificial intelligence products.

Tencent Holdings Ltd., Baidu Inc., JD.Com Inc., Huawei Technologies Co. Ltd., and ByteDance, the parent company of TikTok, are some of Alibaba Cloud’s competitors in the Chinese AI market. All these businesses have introduced competing LLMs in the past 18 months in an attempt to take advantage of the excitement surrounding generative AI technology.

LLMs are AI models that are trained on enormous amount of data to produce responses to user prompts and questions that resemble those of a human. They serve as the basis for a number of generative AI chatbots, including OpenAI’s ChatGPT and Google’s Gemini, as well as picture generators like DALL-E and next-generation search engines like Perplexity AI.

Since Chinese businesses are equally as interested in the potential of generative AI to boost productivity as their American counterparts, Alibaba is concentrating its efforts on them. Alibaba Cloud reported in May that its Qwen models had been downloaded by over 90,000 businesses nationwide.

The business is up against fierce competition. Over 250 new LLMs have been made available to the public by Chinese generative AI developers in the past year. In addition to its largest IT companies, the nation is home to several innovative businesses, like DeepSeek, which just unveiled the DeepSeek-V3 LLM with 671 billion parameters, one of the most potent open-source models in the market.

Qwen-VL, Qwen-VL-Chat, Qwen-VL-Max, Qwen2-VL, and the experimental QVQ-72B-Preview are all included in Alibaba Cloud’s Qwen models. With its cutting-edge performance, Qwen2-VL-Max outperforms Google’s Gemini Ultra and OpenAI’s GPT-4V in benchmarks like DocVQA and MathVista.

The aggressive price reductions used by the Chinese company in an attempt to gain more business are not a novel tactic. It first announced price reductions of up to 55% on several of its key cloud computing services in February. In May, it reported a 97% drop in the price of its original Qwen-VL model.