In the rise of Deepseek, Alibaba launches a new model of artificial intelligence that promises to overcome competitors and intensify the dispute in the technology sector
Chinese technology company Alibaba (9988.HK) launched on Wednesday (29) a new version of its artificial intelligence model, Qwen 2.5, which claims to overcome the highly acclaimed DeepSeek-V3.
According to Reuters, the release of Qwen 2.5-Max occurred at an unusual time-on the first day of the Lunar New Year, when most Chinese is off and reunited with their families. This highlights the pressure imposed by the meteoric rise of the Chinese deepSeek Chinese startup in the last three weeks, not only about international competitors, but also about domestic rivals.
“Qwen 2.5-max surpasses… almost in every respect the GPT-4O, Deepseek-V3 and Llama-3.1-405b,” said Alibaba’s cloud unit in an ad published in his official Wechat account, referring to -the most advanced AI models of OpenAi and the goal.
The launch of DeepSeek’s AI assistant on January 10, fed by the DeepSek-V3 model, as well as the launch of the R1 model on January 20, surprised the Silicon Valley and overthrew technology companies’ actions. The alleged low development and use of Chinese startup costs led investors to question the huge spending planned by the United States’ leading AI companies.
However, Deepseek’s success also triggered a race among its domestic competitors to improve its own AI models.
Two days after the launch of Deepseek-R1, Tiktok’s Bytedance launched an update by its main AI model, claiming that it surpassed OpenAi’s O1 model, supported by Microsoft at AIME-a benchmark test that measures to AI models capacity to understand and respond to complex instructions.
This allegation echoes Deepseek, which stated that its R1 model rivals OpenAi O1 in several performance benchmarks.
Deepseek and the dispute with domestic rivals
The predecessor of the Deepseek-V3 model, the DeepSek-V2, caused a price war between AI models in China after its launch in May last year.
The fact that Deepseek-V2 is open source and has an extremely low cost-only 1 yuan ($ 0.14) by 1 million tokens (data units processed by the AI model)-led the cloud unit of Alibaba to advertise price cuts of up to 97% in a variety of models.
Other Chinese technology companies followed the example, including Baidu (9888.HK), which launched the first chatgpt equivalent in China in March 2023, and Tencent (0700.HK), the most valuable internet company in the country.
Liang Wenfeng, Deepseek’s enigmatic founder, said in a rare interview with the Chinese vehicle Waves in July, that the startup “didn’t care” about the price war and that its main objective was to achieve Agra (General Artificial Intelligence).
OpenAi defines acted as autonomous systems that surpass humans in most economically valuable tasks.
While large Chinese technology companies such as Alibaba have hundreds of thousands of employees, DeepSeek operates as a research laboratory, consisting mainly of young graduates and doctoral students from leading Chinese universities.
Liang said in July that he believes China’s largest technology companies may not be well prepared for the future of the AI industry, contrasting their high costs and strict structures with the lean operation and the flexible management style of DeepSek.
“Fundamental models require continuous innovation, and the skills of technology giants have their limits,” he said.
Source: https://www.ocafezinho.com/2025/01/29/alibaba-desafia-deepseek-com-ia-mais-poderosa/
