BEIJING, Sept 24 (Reuters) – Chinese e-commerce giant Alibaba (9988.HK), opens new tab announced on Wednesday its largest ever artificial intelligence language model, the Qwen3-Max, doubling down on AI as a core business strategy.
The model, Alibaba’s most powerful to date, contains more than 1 trillion parameters, or variables that determine how an AI system processes information, and shows particular strength in code generation and autonomous agent capabilities, Zhou Jingren, chief technology officer at Alibaba Cloud, said at the company’s annual conference.
Autonomous agent capabilities mean the AI system requires fewer human prompts than a chatbot like ChatGPT, and can make decisions and take action independently towards a goal set by the human user.
Alibaba cited third-party benchmarks, such as Tau2-Bench, saying the model outperformed rival products including Anthropic’s Claude and DeepSeek-V3.1 in certain metrics.
Alibaba has made AI a priority alongside its traditional e-commerce operations. Earlier this year, the company announced plans to invest 380 billion yuan ($53.40 billion) in AI-related infrastructure over the next three years as competition to develop advanced AI capabilities intensifies among Chinese tech firms.
During the conference, Alibaba CEO Eddie Wu said the company would increase spending further, though he did not specify the amount.
“The speed of AI industry development has far exceeded our expectations, and the industry’s demand for AI infrastructure has also far exceeded our expectations,” Wu said.
The company released the Qwen 3 model in April.
Alibaba also unveiled several other AI products on Wednesday, including Qwen3-Omni, a multimodal, immersive system useful for virtual and artificial reality applications such as smart glasses and intelligent cockpits.