Alibaba Challenges OpenAI with Its AI Model Qwen3.8-Max

Alibaba Challenges OpenAI with Its AI Model Qwen3.8-Max

Chinese tech giant Alibaba has announced the launch of what it describes as its largest and most powerful AI model to date, claiming that its performance rivals that of the best systems from American labs, such as Anthropic and OpenAI, as well as national competitors like Moonshot AI with its model Kimi K3. Alibaba stated that the model Qwen3.8-Max would be widely available to users in a blog post published on Monday. This release was anticipated after the company showcased the model last month, when it claimed that it was "only second to Fable 5," Anthropic's flagship model.

The release of this high-performance AI model adds to the already heightened tensions at multiple levels in Silicon Valley and Washington regarding the secure management of AI systems and maintaining the technological edge of the United States over China.

Test results conducted by Alibaba, shared on Monday, suggest that the model is as powerful as claimed, as also indicated by its ranking on the model comparison platform Arena.AI. Alibaba's tests show that the model's performance largely matches — and sometimes exceeds — that of Fable 5 on benchmark tests. In the Arena text model rankings, Qwen3.8-Max is only outperformed by Fable 5 and three models from Anthropic's Opus family. In terms of frontend coding, it is surpassed only by two Claude Opus models and Kimi K3, and for visual analysis, only Fable 5 ranks higher.

Alibaba claims that Qwen3.8-Max has 2.4 trillion parameters, a numerical measure of the adjustments a model learns during training and uses to process data, recognize patterns, and perform various tasks. While the number of parameters is commonly used as an indicator of model performance, a higher figure does not always equate to better performance. Moonshot's Kimi K3 has 2.8 trillion parameters, but most leading American labs keep these figures private, and neither OpenAI nor Anthropic disclose the exact counts of their most advanced systems.

The Chinese company announced that it would release the weights of Qwen3.8-Max next week. Weights are the adjustable numerical values of an AI that determine how it processes information. Open-weight systems, while more restrictive than traditional open-source software, offer developers significantly more control than proprietary products from companies like OpenAI and Anthropic. Qwen3.8-Max marks a return to open-weight releases for Alibaba after the company briefly shifted to proprietary releases for its more advanced models earlier this year.

Open-weight releases have become an increasingly differentiating factor for the AI industry in China, where they have become the norm. Moonshot released Kimi K3 last week, and many other leading AI models are also open-weight. Beijing has supported this strategy as a means to increase China's influence in global AI governance and to encourage widespread adoption of national tech champions.

Alibaba's release intensifies competition with China, where companies appear to be rapidly closing the gap with American firms and have accelerated the pace of releases in recent weeks. Qwen3.8-Max closely follows the release of Kimi K3, seen as another challenge to American dominance in AI, and both ByteDance and MiniMax released new high-performing video generation models on Friday.

Openness has also become a topic of debate. Amid reports of a potential crackdown on open tools following Chinese releases, the American industry has largely rallied around preserving access to open-weight models, both as a security necessity and as a means to maintain competition. This debate arises as closed-model providers, including OpenAI and Anthropic, face increased scrutiny after revealing a series of cyberattacks perpetrated unknowingly by their own escaped AI agents. Reports of incidents from one of the victims suggest that restrictive security measures intended to prevent the malicious use of AI models also limit their use as defensive tools.