한국어EN
Open-Weight LineageAlibaba

Alibaba releases Qwen3.8-2.4T-A95B as open weights

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

Qwen3.8-2.4T-A95B released as open weights — a hybrid of Gated DeltaNet and Gated Attention

Why it matters

The top open-weight model reverts to a custom license rather than Apache. Qwen3.8-Max is a closed commercial version built on it

Source — Primary sources first; secondary reports are labelled “Press”

Officialhuggingface.cohttps://huggingface.co/Qwen/Qwen3.8-2.4T-A95B

Added in Week 35, 2026 · See that week

On the map — ⑨ Open-Weight Lineage line

  1. Previous stopK-EXAONE 2.0
  2. This recordQwen3.8-2.4T-A95B
  3. Next stopMistral AI releases the weights of Mistral Large 4 (Le Chonk)
See it on the map

In the model family tree — Qwen family · 13 versions

  1. Earlier versionQwen3.6-27B
  2. This recordQwen3.8-2.4T-A95B
  3. Later versionQwen3.8-27B
See the whole Qwen family

Nearby on the same axis

Alibaba releases Qwen3.8-27BOfficialLG AI Research releases K-EXAONE 2.0 (750B-A37B)OfficialQwen3.8-Max promises open Qwen-Max-class weights, but none are out at launchPressMoonshot AI releases Kimi K3 as open weightsOfficial