Hermes Vector

AI & SITrending

Alibaba Unveils Qwen4-Max AI Model With Safety Upgrades

Alibaba has launched its new flagship AI model, Qwen4-Max, featuring 2.4 trillion parameters and strengthened safeguards against cybersecurity and CBRN risks.

By Hermes Vector

Also in: Español · Português · Français · Deutsch

Checked against 4 sourcesfinance.yahoo.comeconomictimes.indiatimes.comzaobao.com.sgthenews.com.pk
Illustration for: Alibaba Unveils Qwen4-Max AI Model With Safety Upgrades

In brief

  • Alibaba launched Qwen4-Max, a 2.4 trillion parameter AI model with a Mixture-of-Experts architecture.
  • The model activates 71 billion parameters per token, 25% fewer than its predecessor, cutting average task costs by 60%.
  • Alibaba strengthened safeguards against cybercrime, hazardous materials, and chemical, biological, radiological and nuclear risks.
  • Internal testing found no evidence of significant high-risk capabilities, according to the company.

Alibaba Unveils Qwen4-Max AI Model With Safety Upgrades

Alibaba has unveiled its new flagship artificial intelligence model, Qwen4-Max, the company said in reports published on November 15 and 16, 2026.

Model specifications and performance

Qwen4-Max comprises 2.4 trillion parameters and uses a sparse Mixture-of-Experts architecture. It activates 71 billion parameters per token, which is 25% fewer than its predecessor, Qwen3.8-Max, unveiled in August 2026. Alibaba said the model delivers better reasoning capabilities and lowers inference costs, cutting the average cost per task by 60%, according to the company.

The model is available via QwenCloud, priced at 7.5 U.S. cents per 1 million tokens.

Safety upgrades

Alibaba said it enhanced the system's ability to handle complex professional and scientific work while improving protections against safety risks. The company said Qwen4-Max's training included security checks alongside coding tests, helping reduce vulnerabilities such as code injection, memory-safety flaws and broken access controls.

Safeguards to prevent misuse of the model in areas such as cybercrime and hazardous materials were strengthened, while internal testing found no evidence of significant high-risk capabilities, the company added.

The company also introduced dedicated datasets and post-training enhancements to mitigate chemical, biological, radiological and nuclear risks.

Capabilities

Alibaba said, "It can take on a wide range of creative and professional work, spanning 3D modeling, game development, precision manufacturing, finance, law, healthcare, and frontier scientific discovery."

Why it matters

The launch comes as Chinese AI model developers increase safety disclosures in new model releases, amid growing global concern surrounding "rogue AI" and loss-of-control risks. Alibaba's shares were down more than 1% in the overnight session late Sunday, even as the company launched its most advanced AI model.


Sources

This article was drafted with AI assistance and checked against the sources above. Company claims are reported as claims. Cover image is AI-generated.

Questions readers ask

When was the Qwen4-Max AI model unveiled?
Alibaba unveiled the Qwen4-Max model in reports published on November 15 and 16, 2026.
What are the key specifications and performance metrics of Qwen4-Max?
The model comprises 2.4 trillion parameters and activates 71 billion parameters per token, which is 25% fewer than its predecessor. It cuts the average cost per task by 60% and is priced at 7.5 U.S. cents per 1 million tokens via QwenCloud.
What safety upgrades were included in the new model?
Training included security checks to reduce vulnerabilities like code injection and memory-safety flaws, while safeguards were strengthened against misuse in cybercrime and hazardous materials. The company also introduced datasets to mitigate chemical, biological, radiological and nuclear risks.
Why does this launch matter in the current AI landscape?
The launch occurs as Chinese AI developers increase safety disclosures amid global concerns about rogue AI and loss-of-control risks. Despite the release, Alibaba's shares were down more than 1% in the overnight session late Sunday.

The day's checked news, by email

One short email a day with the top stories. Free, and you can unsubscribe in one click.

Or follow for the latest news:

How do you feel about this story?

Comments 0

No comments yet. Start the discussion.

AI & SI

US and nine allies sign San Francisco Declaration to secure AI supply chains

The United States and nine allies signed a declaration to coordinate protection of AI technologies and critical infrastructure, the US State Department said. The signatories include Australia, India, Israel, Japan, South Korea, Qatar, Singapore, the United Arab Emirates and the United Kingdom.

Checked against 2 sourcesforth.newssg.news.yahoo.com