Qwen3.8-Flash-Next-FP4-Local

What is Qwen3.8-Flash-Next-FP4-Local

NVIDIA-released Qwen3.8B Flash Next FP4 quantised version, a lightweight large language model for local execution.

Key features

• FP4 quantised version of Qwen3.8B Flash Next, supporting local execution
• Quantised using Quantisation-Aware Distillation (QAD) technology
• Supports vision encoder, hybrid attention, MoE, gated residual streams and n-gram embeddings
• Provides a vLLM container to support the new NVFP4 lossless compression format

Who it is for

Developers and researchers who need to deploy and run lightweight multimodal large language models on local NVIDIA hardware (such as RTX PRO 6000 Blackwell or DGX Spark).

How to get started

1. Visit https://huggingface.co/nvidia/Qwen3.8-Flash-Next-FP4-Local to download the model weights
2. Run the model using the vLLM container provided in the README
3. Deploy on supported NVIDIA RTX PRO 6000 Blackwell or DGX Spark hardware

Compiled from public descriptions on the tool's official website / README / launch page. Pricing and user numbers have not been verified; please check the official website. Some sections were drafted with AI assistance. Written in Chinese and English; other languages are machine-translated.