release aidev NEEDS REVIEW SIG 1/5

Qwen3.8-27B FP8 quantized model published

Qwen published an FP8 quantized version of Qwen3.8-27B on Hugging Face under the Qwen organisation. The model is image-text-to-text (multimodal), conversational, Apache 2.0 licensed, and derived from the full-precision Qwen3.8-27B base. No capability or benchmark claims are made in the repository listing.

PUBLISHED2026-08-13
OBSERVED2026-08-21
AGE12d
SOURCES1
  • FP8 quantized variant of Qwen3.8-27B (Qwen3.5 generation)
  • Pipeline: image-text-to-text (multimodal)
  • Library: transformers, safetensors format
  • License: Apache 2.0
  • Base model: Qwen/Qwen3.8-27B
  • Tags: endpoints_compatible, region:us
  • No capability, benchmark, or performance data in the source
feature

COMMUNITY

No curated reactions recorded for this event. Facts and takes are kept in separate layers — community context is added by hand, never blended into the record above.