FrontierThe story, in brief
Alibaba debuts Qwen3.8-Max model with 2.4T parameters
2.4T parameters. Alibaba just shipped its most capable model yet—and it's challenging the frontier labs' parameter-scaling playbook.
The KeyNews take
Why it matters
Alibaba's Qwen3.8-Max represents a significant capability jump (7x parameter increase over Qwen3.5) and signals intensifying competition in the frontier model race. The 95B active parameters via mixture-of-experts architecture reflects evolving training strategies to balance scale with efficiency—a capability benchmark the industry watches closely.
The key facts
5 to knowQwen3.8-Max features 2.4 trillion total parameters
Activates 95 billion parameters (likely MoE architecture)
7x larger than Qwen3.5 (released February 2026)
Open-source model from Alibaba
Described as Alibaba's most capable LLM to date
Go to the source
SiliconAnglesiliconangle.com
Publisher excerpt: Alibaba Group Holding Ltd. today debuted a new addition to its Qwen series of open-source large language models. Qwen3.8-Max is the Chinese e-commerce giant’s most capable LLM to date. It features 2.4 trillion parameters, about seven times more than the Qwen3.5 model that Alibaba released in…