Alibaba releases Qwen3.8-Flash-Next, previewing Qwen4 architecture
Qwen3.8-Flash-Next is a multimodal MoE with 125B parameters plus 51B N-gram embeddings, activating only 6B per token. It beats Claude Opus 4.6 Max on 8 of 9 comparable benchmarks. QwenCloud API pricing: $0.16/1M input and $0.47/1M output tokens.
How this story unfolded
7 days · 15 reports · 45 community posts · 60 of 63 shown
- Aug 25
- Aug 26
Alibaba’s Qwen to open-source Qwen3.8-Flash-Next, previewing Qwen4 architecturetechnode.com
unsloth/Qwen3.8-Flash-Next-GGUFhuggingface.co
Alibaba’s Qwen Team Releases Qwen3.8-Flash-Next: A 125B Multimodal MoE With 6B Active Parameters Previewing the Qwen4 Architecturemarktechpost.com
Experiment with Qwen3.8-Flash-Next 176B Model on NVIDIA GB300 NVL72 for Agentic Codingdeveloper.nvidia.com
- Aug 27
- Aug 28
- Aug 29
- Aug 30
- Aug 31
- Sep 1
Qwen by email
Get an email when Qwen has news
No news that day, no email.
More stories today
- Google maps global methane emissions with deep learning
- Sevii expands ADR platform with AI agents for autonomous attack defense
- OpenAI's Prism writing surface gets update from small team
- User asks Claude to draw itself after analyzing chat history
- Claude API blocks editing context before thinking blocks