Chinese technology giant Alibaba has introduced a new artificial intelligence (AI) model, claiming it outperforms competitors such as OpenAI, Meta, and DeepSeek. The announcement of the Qwen2. 5-Max model on January 29 marks the second significant AI revelation from China this week, following DeepSeek's R1 open-weight model, which generated excitement with assertions of superior performance and cost-effectiveness compared to American counterparts. Alibaba asserts that Qwen 2. 5-Max, which is partially open-source, is even more remarkable, outshining several rival models in various internal evaluations conducted by the company. According to Alibaba representatives, in benchmark tests such as Arena-Hard, LiveBench, LiveCodeBench, GPQA-Diamond, and MMLU-Pro, Qwen2. 5-Max matches [Anthropic's] Claude-3. 5-Sonnet and almost entirely exceeds [OpenAI's] GPT-4o, DeepSeek-V3, and [Meta's] Llama-3. 1-405B, as stated on January 28 in a translated message on WeChat. The announcement arrives amidst uncertainty for American tech firms. Following DeepSeek's reveal, their AI chatbot quickly climbed to become the most downloaded free app in Apple's U. S.

App Store. DeepSeek's assertion of superior results while training and operating its model at a significantly lower cost sent shockwaves through the industry, leading to a $1 trillion drop in valuations for major tech companies, including Nvidia, which experienced the largest single-day market loss in U. S. history, valued at $589 billion. The rise of DeepSeek has also sparked a competitive race among China's leading AI companies, prompting TikTok parent company ByteDance to enhance its Doubao model and likely inspiring Alibaba to unveil its own latest model.