China Goes to Fight! Alibaba Unveils Qwen 2.5-Max AI Model Claimed to Outperform OpenAI’s DeepSeek-V3 and GPT-4 in Almost Every Aspect

Chinese tech giant Alibaba has announced the launch of its latest AI model, Qwen 2.5-Max, which it says outperforms the currently popular DeepSeek-V3 model.

“Qwen 2.5-Max outperforms GPT-4o, DeepSeek-V3, and Llama-3.1-405B in almost every aspect,” Alibaba’s cloud division announced on its official WeChat account. The model in question is the most advanced AI model from OpenAI and Meta.

Alibaba’s announcement on Chinese New Year (January 29, 2015), when most Chinese are off work and with their families, reflects the pressure that startup DeepSeek has been under in the past three weeks, affecting not only its foreign competitors but also its domestic competitors.

In addition, TikTok parent company ByteDance has been rushing to release its updated AI model just two days after DeepSeek’s R1 was released, and claims that it outperformed OpenAI’s o1 in an AIME (a measure of the ability to understand and respond to complex instructions) test, demonstrating its “aggressive and urgent competitive strategy.”

DeepSeek founder Liang Wenfeng, who rarely gives interviews, said in an interview with Chinese media Waves in July 2024 that “We are not interested in a price war” and stressed that the main goal is to develop artificial general intelligence (AGI).

Wenfeng believes that Chinese technology companies may not be suitable for competing in the AI ​​industry due to structural and corporate culture constraints.