Alibaba Group Holding previewed its newest flagship artificial intelligence (AI) model, Qwen3.8-Max-Preview, and hyped it as “second only to Fable 5” in performance.
The announcement on Sunday triggered an immediate surge in investor confidence, sending Alibaba’s shares up as much as 5.4% in Hong Kong trading on Monday and positioning it as the top performer on the Hang Seng Index.
The new preview build boasts a massive 2.4 trillion parameters, marking the first time a model within the Qwen family has crossed the trillion-parameter threshold. According to Alibaba, the system is fully multimodal, built to process images, video, and documents alongside standard text.
The company described the architecture as “continuously evolving” and promised to release its weights to the open-source community soon, though specific dates and licensing terms remain unannounced.
During the current introductory trial phase, developers can access the model via Alibaba’s Token Plan subscription, as well as its Qoder and QoderWork platforms, at a promotional rate of just 10% of the standard pricing.
Alibaba’s Qwen team aggressively positioned the model on social media, claiming it ranks “second only to Fable 5,” referring to Anthropic’s recently released Claude Fable 5 frontier model. However, this high-profile claim currently lacks external validation.
Unlike its predecessor, Qwen3.7-Max, which launched in May backed by a comprehensive technical breakdown and a score of 56.6 on the Artificial Analysis Intelligence Index, the Qwen3.8 preview arrived with no model card, no benchmark data, and no activated-parameter count.
Furthermore, independent leaderboards like LMArena have yet to evaluate the system. While Anthropic’s Fable 5 currently commands the top spot on LMArena’s global rankings, Qwen3.7-Max sits considerably lower, meaning enterprise buyers must rely strictly on Alibaba’s self-reported performance metrics for now.
The unveiling comes just three days after Beijing-based startup Moonshot AI launched Kimi K3, a 2.8-trillion-parameter open model that drew substantial industry attention.
These rapid-fire releases signal that Chinese technology firms are successfully scaling up to parameter counts that were previously the exclusive domain of elite U.S. AI labs. Analysts at Goldman Sachs suggest this domestic rivalry will likely trigger an influx of high-end coding and enterprise models, increasing competition across the sector.
To anchor its position as China’s primary AI provider, Alibaba recently secured regulatory clearance to power Apple Intelligence features across iOS, iPadOS, macOS, and visionOS within the country.
To sustain this momentum amid tightening U.S. export restrictions on advanced semiconductors, Alibaba has committed more than 380 billion yuan ($53 billion) to cloud and AI infrastructure over three years. This includes deploying its proprietary Zhenwu M890 AI chip, designed by its T-Head subsidiary, to construct self-reliant domestic computing networks. Until independent researchers can evaluate the promised open weights, however, Qwen3.8’s global rank remains unverified.
