# NVIDIA releases quantized Qwen-Image model on Hugging Face

> The Qwen-Image-NVFP4 model, a 20B‑parameter text‑to‑image diffusion transformer, is now available for download with NVIDIA’s Model Optimizer quantization.

Oossa · 2026-10-12 · https://oossa.com/en/nvidia-releases-quantized-qwen-image-model-on-hugging-face

NVIDIA has posted the Qwen-Image-NVFP4 model on Hugging Face. It is a quantized version of the Qwen‑Image text‑to‑image diffusion transformer, keeping the same architecture while lowering precision for faster GPU inference. The model supports up to 1,024 prompt tokens and runs on NVIDIA Blackwell GPUs with TensorRT‑LLM VisualGen. It is released under the Apache 2.0 license for commercial and non‑commercial use.

## The facts

- Release date: 2026-10-12 ("Release Date: Hugging Face 10/12/2026")
- Model size: 20.43 B parameters in the denoising transformer (28.85 B total including text encoder and VAE)

## Why it matters

Developers can now generate images faster on NVIDIA GPUs without needing to quantize the model themselves.

## Sources & references

1. [New model on Hugging Face: nvidia/Qwen-Image-NVFP4 (text-to-image)](https://huggingface.co/nvidia/Qwen-Image-NVFP4) – NVIDIA, 2026-10-09

Last updated: 2026-10-12