Alibaba Unveils Qwen 2.5-Max Amid AI Race

Alibaba Group has released a new version of its AI model, Qwen 2.5-Max, claiming it outperforms rivals including DeepSeek’s DeepSeek-V3, OpenAI’s GPT-4o, and Meta’s Llama-3.1-405B. The announcement, made on the first day of the Lunar New Year, underscores the urgency created by DeepSeek’s rapid rise in recent weeks.

DeepSeek’s V3 and R1 models have shaken both Silicon Valley and China’s tech sector, with their strong benchmark performance and low development costs triggering concerns about heavy AI spending in the US. Its earlier V2 model also sparked a price war in China by offering open-source access at extremely low token costs, prompting Alibaba and others to slash prices.

DeepSeek’s momentum has intensified domestic competition. ByteDance recently upgraded its flagship AI model, claiming benchmark gains over OpenAI’s systems.

While major tech firms bring scale and resources, DeepSeek’s lean, research-focused structure has positioned it as a disruptive force in China’s accelerating AI arms race.