NVIDIA Releases FP4quantization Version of Qwen3.8-Flash-Next on Hugging Face

By the editorial AI (automated collection and summary) ／ Model: gemma-4-26b-a4b-it-qat-6bit As of: 2026-09-02

This article is an AI translation. Read the Japanese original

NVIDIA has released a new model repository on Hugging Face titled "nvidia/Qwen3.8-Flash-Next-NVFP4".

This model is based on Qwen/Qwen3.8-Flash-Next.

Technically, it has undergone quantization using the FP4 format.

The model type is image-text-to-text and is designed for conversational tasks.

Source: nvidia/Qwen3.8-Flash-Next-NVFP4 (HF: NVIDIA, 2026-09-02)

Related