 |
| At the World Financial Intelligence Conference (WIC) in Shanghai, China, on July 6, 2023, an Alibaba signboard is visible | Photo: Reuters |
Chinese tech giant Alibaba has unveiled its latest AI model the Kuen 2.5-Max claiming it surpasses popular models like DeepSeek V3 in terms of performance. The announcement was made on January 29, coinciding with the first day of the Chinese New Year, signaling a major leap in Alibaba's AI capabilities.
The launch came at a time when DeepSeek's user base had been steadily growing over the past three weeks, creating significant competition not only in the international tech market but also within China. The unveiling of Alibaba's Kuen 2.5-Max has intensified this internal competition as the company aims to establish its AI supremacy.
In a statement posted on their official WeChat account, Alibaba's Cloud Unit highlighted the advancements of their new AI model, stating that the Kuen 2.5-Max outperforms OpenAI's GPT-4, DeepSeek V3, and Meta's LLaMA 3.1-405B across almost all areas.
The company also mentioned the recent releases from DeepSeek, including their AI assistant based on the DeepSeek V3 model, launched on January 10, and the R-One model on January 20. These releases stunned Silicon Valley and caused a drop in technology stocks. In just three weeks, they raised questions about the future of AI investments from top US companies.
As DeepSeek rapidly rose to prominence, Chinese tech companies began to intensify their efforts to compete. Several firms have started updating their AI models, responding to the increasing pressure to innovate.
Just two days after the release of DeepSeek's R-One model, ByteDance, the parent company of TikTok, launched its own advanced AI model on January 22. ByteDance claims its AI model is more effective than OpenAI's O1 model, further adding to the growing competition in the field of AI technology.