China has made another decisive move in the global artificial intelligence race, with technology heavyweight Alibaba and its backed startup Moonshot AI unveiling new flagship AI models aimed at competing directly with industry leaders OpenAI and Google. The companies claim their latest systems match global peers in reasoning, coding and autonomous task execution, signalling China’s accelerating push to narrow the AI gap with the United States.
Alibaba Cloud has introduced Qwen3-Max-Thinking, its most advanced and largest AI model to date. The company says the model has been designed to handle complex, multi-step reasoning, tool usage and real-world problem solving more efficiently than previous versions. Researchers involved in its development say the model shows significant improvements in so-called agentic capabilities, allowing it to plan, adapt and execute tasks with minimal human intervention.
Certified Cyber Crime Investigator Course Launched by Centre for Police Technology
According to Alibaba researchers, Qwen3-Max-Thinking is built on more than one trillion parameters, placing it among the largest language models currently in development. In the AI industry, higher parameter counts are often associated with greater learning capacity and more nuanced understanding. Internal evaluations and early benchmarks suggest the model performs competitively against leading systems such as Google’s Gemini and Anthropic’s Claude across several reasoning and coding tasks.
Company executives say the model reflects Alibaba’s long-term strategy of integrating AI more deeply into cloud services, enterprise solutions and consumer applications. The focus, they add, is not only on raw scale but on making AI systems more practical and reliable in real-world scenarios.
Alongside Alibaba’s announcement, Moonshot AI, a Beijing-based startup backed by Alibaba, has launched its new open-source model Kimi K2.5. The company is positioning it as one of the most powerful open-source AI models globally, targeting developers and enterprises seeking flexible, high-performance systems without proprietary restrictions.
One of Kimi K2.5’s standout features is its “agent swarm” capability. This allows up to 100 smaller AI agents to work simultaneously on a single task, significantly speeding up complex operations such as large-scale coding, data analysis and automated workflows. Developers say such parallel processing could reduce development time and improve efficiency for advanced software projects.
Moonshot AI has also expanded Kimi’s scope beyond text. The model now supports image and video processing, marking a shift toward fully multimodal AI systems that can analyse and interpret different types of content within a single framework. Industry observers see this as a crucial step, as leading global models increasingly move toward multimodal functionality.
Despite the technological progress, challenges remain. As AI models grow larger and more sophisticated, the demand for computing power, advanced chips and energy continues to rise. A senior member of the Qwen development team has acknowledged that maintaining a balance between cutting-edge research and stable commercial services is becoming increasingly difficult. High-performance semiconductors, data-centre capacity and operating costs are emerging as major constraints.
Experts note that Chinese AI firms are already looking ahead to models with 10 trillion parameters, which could represent a new leap in capability. However, achieving that scale will not be easy. US export controls on advanced semiconductors and limited access to top-tier chip manufacturing technology could slow progress, even as domestic alternatives improve.
Even so, the latest launches underline a broader shift in the global AI landscape. China’s AI ecosystem is no longer confined to following Western innovation cycles; instead, it is increasingly setting its own pace and priorities. Analysts say the success of Qwen3-Max-Thinking and Kimi K2.5 will ultimately depend on how well they perform in large-scale, real-world applications and how widely they are adopted by developers and businesses.
For now, Alibaba and Moonshot’s announcements make one point clear: the battle for AI leadership is intensifying, and the race is no longer dominated by a handful of Western players. The next phase of competition is set to be faster, broader and far more global.
About the author – Rehan Khan is a law student and legal journalist with a keen interest in cybercrime, digital fraud, and emerging technology laws. He writes on the intersection of law, cybersecurity, and online safety, focusing on developments that impact individuals and institutions in India.
