Alibaba claims its AI model Qwen 2.5-Max surpasses DeepSeek-V3
![Alibaba claims its AI model Qwen 2.5-Max surpasses DeepSeek-V3](https://cdn.jagonews24.com/media/imgAllNew/BG/2023March/qwen25-max-20250129135028.jpg)
Chinese tech giant Alibaba (9988.HK) unveiled its latest artificial intelligence model, Qwen 2.5-Max, on Wednesday, claiming it outperforms the much-hyped DeepSeek-V3 and other leading AI models such as OpenAI’s GPT-4o and Meta’s Llama-3.1-405B.
The release came on the first day of the Lunar New Year, a time when most Chinese citizens are celebrating with their families, signaling the mounting pressure within the AI industry. The meteoric rise of Chinese AI startup DeepSeek over the past three weeks has shaken not only international rivals but also domestic players like Alibaba.
“Qwen 2.5-Max outperforms ... almost across the board GPT-4o, DeepSeek-V3, and Llama-3.1-405B,” Alibaba’s cloud unit announced on its official WeChat account.
DeepSeek's rapid ascent stirs global and domestic markets
DeepSeek's success has taken the tech world by storm. The January 10 release of its DeepSeek-V3-powered AI assistant, followed by the January 20 launch of its R1 model, caused ripples in Silicon Valley, leading to a drop in tech shares. Investors are now questioning the enormous budgets of U.S.-based AI firms, as DeepSeek's models boast significantly lower development and usage costs.
The competitive landscape within China has also intensified. Just two days after DeepSeek-R1’s launch, ByteDance, the parent company of TikTok, rolled out an updated AI model, claiming it outperformed OpenAI’s o1 on AIME, a benchmark for evaluating AI performance in understanding and executing complex instructions.
DeepSeek's earlier success, particularly with its DeepSeek-V2 model, launched in May 2023, triggered a fierce price war in China’s AI market. The model, which was open-source and priced at just 1 yuan ($0.14) per 1 million tokens, forced Alibaba to slash prices on its AI models by up to 97%. Major competitors, including Baidu and Tencent, also responded with their own updates to stay competitive.
DeepSeek's Vision: Innovation over cost wars
DeepSeek’s enigmatic founder, Liang Wenfeng, has made it clear that the company is not focused on price wars but on achieving AGI (artificial general intelligence)—systems that surpass humans in most economically valuable tasks.
In a rare interview with the Chinese media outlet Waves in July 2023, Liang contrasted DeepSeek’s lean and research-focused structure with the high costs and hierarchical management of China’s tech giants.
“Large foundational models require continued innovation, and tech giants have their limits,” Liang remarked. He believes the top-down organisational models of companies like Alibaba and Tencent may hinder their ability to adapt to the fast-evolving AI landscape.
While Alibaba and other established players rely on their vast resources and workforce, DeepSeek operates more like a research lab, staffed primarily by young graduates and PhDs from China’s leading universities.
Future implications
As Alibaba's Qwen 2.5-Max enters the fray, the competition between established tech giants and lean startups like DeepSeek continues to shape the future of AI innovation in China. Whether Qwen 2.5-Max can live up to Alibaba’s claims remains to be seen, but the race for dominance in AI—and potentially AGI—is clearly intensifying.