unsloth / text-to-image Open weights updated 2 weeks ago

Qwen-Image-2.1-FP8

Name: Qwen-Image-2.1-FP8
Author: unsloth

This is an FP8 / INT8 quantized version of Qwen-Image-2.1, plus a pre-cast FP8 copy of its Qwen3-VL text encoder. - For higher quality, you can also run Dynamic FP8 on just 6GB of VRAM via offloading (

Params

7.1 B

Context

—

Downloads 30d

419 K

Likes

Licence: other Not gated SAFETENSORS 2 languages View on Hugging Face ↗

Download history

daily snapshots · 29 days

419 K27 K

Sep 26Oct 5Oct 15Oct 24

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	fp16	41.1 GB	46.7 GB	❌ Won’t fit

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Parameters: 7.1 B
Licence: other
First seen on the Hub: 2026-09-21
Base model: Qwen-Image-2.1
Training datasets: undisclosed
Added to our catalog: 2026-09-26

Family

Base model and the most-downloaded derivatives in the catalog.

Base Qwen-Image-2.1 Qwen 171 K

Compare with any text-to-image model

Popular comparisons

stable-diffusion-xl-base-1.0 4.1 M Qwen-Image-2.1-Uncensored-Quantized 2.4 M Qwen-Image-2.1-Uncensored-GGUF 2.2 M Qwen-Image-2.1-GGUF 1.5 M Z-Image-Turbo-GGUF 1.1 M Z-Image-Turbo 651 K