Chinese tech company Alibaba has released a new version of its artificial intelligence model, Qwen 2.5, saying it performs better than DeepSeek-V3, one of the top AI models in the industry.
Alibaba made the announcement on Wednesday, the first day of the Lunar New Year, a time when most Chinese people are on holiday. The timing suggests the company is under pressure due to the rapid rise of DeepSeek, a Chinese AI startup that has gained global attention in recent weeks.
“Qwen 2.5-Max outperforms … almost across the board GPT-4o, DeepSeek-V3, and Llama-3.1-405B,” Alibaba’s cloud unit said in a statement on its official WeChat account. It compared its new AI model to OpenAI’s GPT-4o and Meta’s Llama-3.1-405B.
DeepSeek has been making headlines since launching its DeepSeek-V3 AI model on January 10 and its R1 model on January 20. The company’s rapid success has shocked Silicon Valley, causing tech shares to fall. Investors have started questioning the high costs of AI development by major U.S. companies, as DeepSeek claims to offer similar capabilities at a lower cost.
DeepSeek’s rise has also led to competition among Chinese AI companies. Just two days after the release of DeepSeek-R1, TikTok owner ByteDance updated its AI model, claiming it outperformed OpenAI’s o1 in AIME, a test that measures how well AI models understand and respond to complex instructions.
The competition started in May 2024 when DeepSeek released its V2 model, which led to a price war in the AI industry. DeepSeek-V2 was open-source and extremely cheap, costing only 1 yuan ($0.14) per 1 million tokens—units of data processed by AI. In response, Alibaba’s cloud unit cut its AI model prices by up to 97%. Other companies like Baidu and Tencent also joined the price cuts.
Despite this, DeepSeek’s founder, Liang Wenfeng, has said the company is not focused on price wars but on achieving artificial general intelligence (AGI), where AI systems surpass human abilities in most tasks.
“Large foundational models require continued innovation, and tech giants’ capabilities have their limits,” Liang said in an interview last year. He believes large Chinese companies like Alibaba may not be suited for the future of AI because of their high costs and rigid structures.
Unlike Alibaba, which has hundreds of thousands of employees, DeepSeek operates like a research lab, with a small team of young graduates and doctorate students from top Chinese universities.
With the AI race heating up, Alibaba and other tech giants are working to stay ahead, while DeepSeek continues to challenge the industry with its low-cost and high-performance AI models.