← Back to the wire

Qwen/Qwen3.8-27B · Hugging Face

AnnouncementModelAug 13, 2026

Hugging Face hosts Qwen/Qwen3.8-27B, a dense multimodal model that processes images and videos with flexible thinking control. The repository provides model weights compatible with Transformers, vLLM, SGLang, and Docker Model Runner. Qwen Cloud plans to offer a hosted version with a 1M context length and built-in tools. Built on the Qwen3.5 architecture, the model targets improvements in coding, research, and agentic task execution.

Receipt № 14181 source · awaiting confirmation ◐

Evidence

1source· awaiting independent confirmation

No score is assigned. Sources and their independence are shown in the citation chain below.

Citation chain · 1 source

01highPRIMARY
Hugging FaceCompanyQwen/Qwen3.8-27BModel
Canonical: https://huggingface.co/Qwen/Qwen3.8-27B