Alibaba Unveils Qwen-3-Max-Preview, Its Largest AI Model Yet
Alibaba launches Qwen-3-Max-Preview, a text-only AI model with over one trillion parameters. The system is available through Alibaba Cloud and represents a major step in the company's $52 billion AI investment strategy.
Alibaba releases Qwen-3-Max-Preview, its most powerful artificial intelligence model to date, featuring over one trillion parameters. This text-only system becomes available through Alibaba Cloud and the OpenRouter platform, marking a significant milestone in the company's AI development. Unlike some competing models, this release remains closed-source with access strictly limited to official channels to control its application.
The massive model handles very long text sequences and tackles a wide range of tasks including writing, summarizing, and answering questions. Alibaba claims the system outperforms previous versions and many rival models, though independent verification of these internal benchmarks is still pending. The architecture also optimizes computational resources to maintain faster response times despite the enormous scale.
This launch aligns with Alibaba's massive $52 billion investment plan for AI infrastructure, research, and development over the next three years. Users access the model through Model Studio at a cost of $0.9 per million input tokens and $3.4 per million output tokens, highlighting the company's strategy to monetize its expanding AI capabilities.