Alibaba Releases Qwen4-Max with 2.4T Sparse MoE and 1M Context

Written by

in

Alibaba has released Qwen4-Max, whose 2.4T sparse MoE backbone activates 71B parameters per token. It can process text, images, and video, and supports a 1M-token context.

Source:MarkTechPost · Read Original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *