AI Summary10 Sept 202629 Rabiʻ I 1448 AH
AI Summary
10 Sept 202629 Rabiʻ I 1448 AH
Alibaba Opens Qwen3.8-2.4T-A95B Weights as First Qwen-Max-Class MoE
Alibaba's Qwen team has announced the release of open weights for the Qwen3.8-2.4T-A95B model, marking the first Qwen-Max-class model available for download instead of API access.
The model features approximately 2.4 trillion parameters, activating about 95 billion per token across 512 experts, with 10 routed experts and one shared expert used at each step. This release positions a frontier-scale text MoE, making it suitable for self-hosted AI applications.
Follow these topics
Sign in to follow the topics that matter to you
This summary is generated with AI and receives periodic editorial review. Refer to the original source for full details.
Alibaba Opens Qwen3.8-2.4T-A95B Weights as First Qwen-Max-Class MoE
Alibaba's Qwen team has announced the release of open weights for the Qwen3.8-2.4T-A95B model, marking the first Qwen-Max-class model available for download instead of API access.
The model features approximately 2.4 trillion parameters, activating about 95 billion per token across 512 experts, with 10 routed experts and one shared expert used at each step. This release positions a frontier-scale text MoE, making it suitable for self-hosted AI applications.
Models
Follow these topics
Sign in to follow the topics that matter to you
This summary is generated with AI and receives periodic editorial review. Refer to the original source for full details.
0
0 reading now
Insight Score
Rate to unlock
You have not rated yet.
Sign in to react, rate, and save. Sign In
