Alibaba unveils Qwen3.8-Max, a 2.4T-parameter MoE with open weights to follow
Alibaba unveiled Qwen3.8-Max, a sparse mixture-of-experts model with 2.4 trillion total and about 95B active parameters aimed at coding and 'cowork' agent tasks, and committed to releasing open weights. Weights for Qwen3.8-2.4T-A95B, without image input, and a smaller Apache-licensed 27B model followed later in August, with commercial terms for large providers. For practitioners, it was Qwen's most capable open release, though its license terms need careful review before commercial use.
Read the full story at Alibaba Qwen ↗Like, save or discuss on Buzz
