AIHacker Newsabout 2 hours ago
Qwen/Qwen3.8-2.4T-A95B
1 min read
Qwen released Qwen3.8-2.4T-A95B, a 2.4T-parameter MoE model with 95B activated, natively supporting 262,144 tokens and extensible to 1,010,000. The open weights are compatible with vLLM, SGLang, and TokenSpeed. Qwen3.8-Max, the official API version, adds vision input, non-thinking support, and built-in tools.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
