Alibaba has unveiled Qwen4-Max, its new flagship artificial intelligence model, saying it handles complex professional and scientific tasks better than its predecessor while adding stronger protections against safety risks, according to The Economic Times. The company announced the model on Sunday.
The release comes as Chinese technology firms race to build more capable AI systems without making them too expensive to run, a contest the outlet described as fierce and fast-moving.
Alibaba presented Qwen4-Max as its top model, and said it improves the system's ability to take on demanding professional and scientific work. The company also said it strengthened defenses against safety problems. The Economic Times reported that the model outperforms Qwen3.8-Max, which Alibaba released in August of this year.
Alibaba said the model can handle a wide range of creative and professional jobs. The company listed 3D modeling, game development, precision manufacturing, finance, law and healthcare. It also named frontier scientific discovery as an area of use. These are the company's own descriptions of what the model can do.
The company said security checks were built into training alongside coding tests. According to Alibaba, that approach helped reduce certain software weaknesses, including code injection, memory-safety flaws and broken access controls.
Alibaba also said it reinforced safeguards meant to stop the model from being misused in areas such as cybercrime and hazardous materials. In its internal testing, the company said, it found no evidence that the model has significant high-risk capabilities.
Alibaba said Qwen4-Max beats Qwen3.8-Max while using fewer tokens, which it linked to better cost efficiency. Tokens are the units of text that AI systems process. The Economic Times noted that Chinese companies, a major force in open-weight AI models worldwide, are trying to build more powerful systems without pushing running costs too high. Alibaba did not disclose pricing in the reporting.
For readers who use or follow AI tools, the announcement shows where the competition is heading. Developers are promoting efficiency and safety alongside raw capability. A model that does more with fewer tokens could, in principle, be cheaper to operate, though Alibaba has not published figures in the reporting.
The safety claims also reflect a wider trend. The Economic Times reported that Chinese AI developers have in recent months published more safety information with new model releases, as global concern grows over so-called rogue AI and the risk of losing control of advanced systems.
Readers should note that the performance comparison and the safety findings come from Alibaba itself. The internal testing result, that no significant high-risk capabilities were found, is the company's own conclusion. The reporting does not describe independent evaluation.
The reporting does not detail how and when businesses or the public can use Qwen4-Max, or what it will cost. Those details were not disclosed.
Around Prague, based on reporting by The Economic Times