Photo: Tingshu Wang/Reuters
Chinese tech corporation Alibaba on Wednesday officially unveiled an updated version of its Qwen 2.5 artificial intelligence, which the company claims is superior to the widely popular DeepSeek-V3 model.
The presentation of Qwen 2.5-Max took place at an unusual time – on the first day of the Chinese New Year, when most citizens are on holiday with their families. This is a sign of the intense competition in the AI market, which has been driven by the rapid success of DeepSeek. Over the past three weeks, its breakthrough models have put pressure not only on international competitors but also on companies within China, Reuters reports.
An official post by Alibaba’s cloud division on WeChat reads: “Qwen 2.5-Max outperforms GPT-4o, DeepSeek-V3, and Llama-3.1-405B in almost every way.”
The unexpected success of DeepSeek, which unveiled DeepSeek-V3 on 10 January and the R1 model on 20 January, caused fluctuations in the stock markets, which led to a drop in the value of shares of tech giants, including Nvidia, which lost $600 billion in capitalisation.
Analysts point out that investors are concerned about the low costs of developing and operating DeepSeek models, which calls into question the feasibility of large-scale investments by American AI companies. In response to the rapid development of DeepSeek, competitors in China have also started to update their AI models.
In particular, two days after the launch of DeepSeek-R1, ByteDance (owner of TikTok) presented an improved version of its AI. It stated that its model outperformed OpenAI’s GPT-4o in the AIME test, which assesses the ability of models to understand and execute complex instructions.
