BEIJING, Aug 4 — Chinese technology and e-commerce giant Alibaba has unveiled another powerful artificial intelligence model as the global race for AI leadership intensifies. Its new Qwen3.8-Max model competes with the leading systems from US rivals OpenAI and Anthropic, reported German Press Agency (dpa).
Alibaba on Monday described Qwen3.8-Max as its largest and most capable AI model to date. The company said it excels at programming, conducting deep research and handling complex tasks with a high degree of autonomy.
The model has 2.4 trillion parameters, a commonly used measure of an AI system’s size and complexity. A higher parameter count, however, does not necessarily translate into better performance.
Alibaba said Qwen3.8-Max outperformed the recently launched Kimi K3 from Chinese developer Moonshot AI in several benchmark tests, despite Kimi K3 having 2.8 trillion parameters.
On Arena.AI, a popular AI benchmarking platform, Qwen3.8-Max ranked as the highest-performing Chinese model for text-based tasks, although it trailed models from US-based Anthropic. It performed particularly strongly in image analysis, Alibaba said. The model is already available for use.
The rapid succession of new model launches underscores the fierce competition within China’s AI industry while increasing pressure on leading US developers.
Chinese AI systems are increasingly matching top US models in some tasks, and many Chinese companies are releasing their model weights — key information about the AI system — to allow developers to customise and fine-tune the systems themselves. — Bernama-dpa
