Alibaba unveils AI model that rivals DeepSeek

Alibaba has released a new AI model, Qwen 2.5-Max, that it claims surpasses DeepSeek’s V3 model. This announcement comes on the first day of the Lunar New Year, a time when most Chinese people are at home with their families, suggesting the intense pressure that DeepSeek’s rapid growth has put on its competitors, both domestically and internationally.

According to Alibaba’s cloud unit, Qwen 2.5-Max “outperforms almost all GPT-4o, DeepSeek-V3 and Llama-3.1-405B,” referring to the advanced AI models from OpenAI and Meta, as stated in an announcement on their official WeChat account. The launch of DeepSeek’s AI assistant, powered by the DeepSeek-V3 model on January 10 and the R1 model on January 20, caused shockwaves in Silicon Valley and a downturn in tech stocks. Investors began questioning the expensive strategies of leading U.S. AI companies due to DeepSeek’s low development and usage costs.

The rise of DeepSeek has also intensified competition among its domestic rivals, leading them to upgrade their AI models. Just two days after the release of DeepSeek-R1, ByteDance, the owner of TikTok, launched an updated version of its flagship AI model, claiming it outperformed Microsoft-backed OpenAI’s o1 model on the AIME benchmark test, which evaluates AI models’ ability to understand and respond to complex instructions. This reinforced DeepSeek’s assertion that its R1 model was comparable to OpenAI’s o1 across various performance metrics.

The predecessor to DeepSeek’s V3 model, DeepSeek-V2, triggered an AI model price war in China following its release last May. DeepSeek-V2’s open-source nature and extremely low price of 1 yuan ($0.14) per 1 million tokens prompted Alibaba’s cloud unit to significantly reduce prices on a range of its models, with cuts reaching up to 97 percent. Other Chinese tech companies such as Baidu, which launched China’s first ChatGPT equivalent in March 2023, and Tencent, the country’s most valuable internet company, followed suit.

In a rare interview with the Chinese media outlet Waves in July, DeepSeek’s founder, Liang Wenfeng, stated that the startup “doesn’t care” about price wars and that its primary focus is achieving AGI (artificial general intelligence). OpenAI defines AGI as autonomous systems that surpass human capabilities in most economically significant tasks. Unlike large Chinese tech companies like Alibaba, which employ hundreds of thousands of people, DeepSeek operates more like a research lab, primarily staffed by young graduates and PhD students from top Chinese universities.

In an interview in July, Liang Wenfeng suggested that China’s largest tech companies may not be well-suited for the future of the AI industry due to their high operational costs and rigid structures, contrasting them with DeepSeek’s streamlined operations and flexible management style. He noted that “large basic models require constant innovation, while the capabilities of tech giants have limits.”