AIThe Next Webabout 2 hours ago
Qwen4 architecture preview fires 6B of 125B
1 min read
Alibaba's Qwen team released Qwen3.8-Flash-Next, an open-weight preview of the Qwen4 architecture with 125B parameters but only 6B active per token. The model adds 51B embedding parameters and may not qualify for the EU AI Act's open-source exemption.
Level
Hype check
Tap to vote and see what everyone thinks.
