Alibaba Group (NYSE:BABA) has launched its new artificial intelligence model, Qwen2.5-Max, claiming it surpasses leading models such as OpenAI's GPT-4o and DeepSeek-V3. The model is now available on the Alibaba Cloud Bailian platform, accessible to developers through an interactive chat tool.
According to Alibaba, Qwen2.5-Max outperforms DeepSeek-V3 on key benchmarks, including Arena-Hard, LiveBench, LiveCodeBench, and MMLU-Pro. The model is also said to be competitive with Anthropic's Claude-3.5-Sonnet.
This release follows the introduction of DeepSeek's R1 model, which has raised concerns regarding the sustainability of large-scale AI infrastructure investments. The rise of cost-effective Chinese AI models has contributed to a $1 trillion sell-off in U.S. tech stocks, particularly affecting Nvidia. Investors are now questioning the long-term demand for high-end AI chips from major firms like Microsoft and Meta.