Print Page - Qwen3.8-27B released! (Qwen compares it again Opus4.6 Max)

Title: Qwen3.8-27B released! (Qwen compares it again Opus4.6 Max)
Post by: third eye on August 14, 2026, 17:58:41

huggingface.co/Qwen/Qwen3.8-27B

Crazy benchmark scores, if not benchmaxxed.

Several posts and discussion on reddit.com/r/LocalLLaMA.

Most will be interested in running the quants, not the full weights: unsloth.ai/docs/models/qwen3.8#quantization-analysis

artificialanalysis.ai will evaluate it soon.

Title: Re: Qwen3.8-27B released! (Qwen compares it again Opus4.6 Max)
Post by: third eye on August 14, 2026, 18:04:00

against*

NotebookCHECK - Notebook Forum

English => News => Topic started by: third eye on August 14, 2026, 17:58:41