Chinese tech and e-commerce giant Alibaba announced the release of Qwen 2.5-Max on Wednesday, an advanced AI model that the company claims outperforms several leading systems in key benchmarks.

To advertise here,

This launch follows the recent release of models by Chinese startup DeepSeek, which have stunned Silicon Valley and challenged the long-standing assumption of U.S. dominance in the AI sector.

The emergence of consecutive high-performing Chinese models is likely to heighten concerns in the United States, where companies have invested billions in AI development, only to see Chinese startups deliver competitive results at a fraction of the cost.

In a blog post, the Qwen team highlighted that their new model outperformed DeepSeek V3 in several tests, including code generation and general capabilities, and showed competitive results against industry leaders like OpenAI's GPT-4 and Anthropic's Claude-3.5-Sonnet.

Qwen 2.5-Max was trained on over 20 trillion tokens of data, but it was notably not compared to DeepSeek's R1 model, which is a reasoning-focused AI that gained attention when it was launched as a chatbot on January 20.

Now available to developers via Alibaba Cloud services, Qwen 2.5-Max can be accessed through Qwen Chat, Alibaba's conversational AI platform. The system is compatible with OpenAI's API format, potentially making it easier for organizations already using similar AI services to adopt the model.