Chinese tech giant Alibaba on Wednesday introduced an upgraded version of its Qwen 2.5 artificial intelligence model, asserting that it surpasses the highly acclaimed DeepSeek-V3, according to Reuters.
The unexpected launch of the AI model on the first day of the Lunar New Year—a time when most Chinese people are on holiday—highlights the intense competition in the industry following the rapid rise of Chinese AI startup DeepSeek. The newcomer has shaken both domestic and international rivals with its breakthroughs.
“Qwen 2.5-Max outperforms GPT-4o, DeepSeek-V3, and Llama-3.1-405B across almost all benchmarks,” Alibaba’s cloud unit stated in an official WeChat announcement, referencing OpenAI and Meta’s most advanced open-source AI models.
DeepSeek’s Disruptive Rise in AI
On January 10, DeepSeek introduced its AI assistant powered by DeepSeek-V3, and just ten days later, it launched the R1 model—a low-cost, high-performance AI system that sent shockwaves through the industry.
Merely two days after the release of DeepSeek-R1, TikTok’s parent company, ByteDance, rolled out an updated version of its flagship AI model, claiming it had surpassed Microsoft-backed OpenAI’s o1 in AIME—a benchmark evaluating AI comprehension and responsiveness to complex instructions.
DeepSeek’s disruptive impact was already evident in May 2024, when the launch of its DeepSeek-V2 model triggered a price war in China’s AI market. In response, Alibaba’s cloud division slashed AI model prices by up to 97%, prompting other companies to follow suit.
Despite these market shifts, DeepSeek founder Liang Wenfeng emphasized in a July interview with Chinese media outlet Waves that his company was unfazed by pricing battles, stating that achieving artificial general intelligence (AGI) remains its ultimate goal.
