ToplineChinese tech giant Alibaba on Monday launched what it called its “most capable” AI model yet, which the tech giant claimed tested comparable to or better than Anthropic’s Fable and OpenAI’s ChatGPT amid China’s accelerating AI race with the U.S.The Chinese tech giant hailed its model as its “most capable.” Associated PressKey FactsAlibaba introduced its Qwen3.8-Max AI model, scheduled to release next week, as one of the world’s largest open AI models at 2.4 trillion parameters, or the values that shape how a model processes information.The model’s capabilities include real-life work, research, autonomous coding and long-horizon execution, an AI model’s ability to complete tasks that require many sequential steps, Alibaba said.Qwen3.8-Max can code for hours without user input, and in one internal test, Alibaba claimed the model spent 16 days building and improving an AI coding tool by writing code, testing it, fixing errors and then refining its work on its own, which Alibaba said showed improved efficiency over time. Alibaba shared testing results that appeared to show Qwen3.8-Max outperforming Anthropic’s Fable 5 in multimodal reasoning, visual agent and coding, office intelligence, real-world understanding, and visual perception, with results comparable to or higher than OpenAI’s GPT5.6-Sol.Shares of Alibaba jumped 4.8% shortly after trading opened on Monday, following a 7% rise in Hong Kong-based trading.surprising factAlibaba’s AI model is comparable in size to Chinese startup Moonshot AI’s Kimi K3, which the firm claimed was the largest-ever open AI model at 2.8 trillion parameters. Open models, like Kimi K3 or Qwen3.8-Max, allow developers to download, run and modify the AI, unlike proprietary models like ChatGPT and Anthropic’s Claude, whose underlying systems are private. Neither OpenAI nor Anthropic has disclosed the total parameter counts of their models. key backgroundAhead of the full unveiling of Qwen3.8-Max, Alibaba claimed its AI model would be comparable to other leading AI models and second only to Anthropic’s flagship Fable. The release comes as Chinese AI firms have worked to close the gap with their American competitors, despite U.S. export controls limiting access to the most advanced AI chips, including Nvidia’s H200 chips. The U.S. has alleged that Chinese firms have smuggled banned AI chips into China from the U.S. and worked around export restrictions by buying processors through foreign subsidiaries. Last month, Treasury Secretary Scott Bessent said the Trump administration would look into whether Chinese AI models had been distilled from American models, shortly after Moonshot—founded in 2023 with backing from Alibaba—launched its Kimi K3 model, adding the U.S. government did not support “IP theft.”further readingForbesChinese AI Startup Moonshot Unveils Kimi K3 Model—Will It Challenge OpenAI And Anthropic?By Ty Roush