unsloth / text-to-image Open weights updated 1 week ago

Z-Image-GGUF

Name: Z-Image-GGUF
Author: unsloth

This is a GGUF quantized version of Z-Image. unsloth/Z-Image-GGUF uses Unsloth Dynamic 2.0 methodology for SOTA performance. - Important layers are upcasted to higher precision. - Uses tooling from ComfyUI-GGUF by city96.

Params

6.2 B

Context

—

Downloads 30d

224 K

Likes

204

Commercial use: allowed apache-2.0 Not gated GGUF · SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 82 days

▲ 100 K in the last 30 days (80.7%)

224 K18 K

Aug 21Sep 17Oct 14Nov 10

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
z-image-Q2_K.gguf	Q2_K	2.3 GB	4.0 GB	✅ Runs comfortably
z-image-Q3_K_M.gguf	Q3_K_M	3.3 GB	5.1 GB	✅ Runs comfortably
z-image-Q4_K_M.gguf	Q4_K_M	3.9 GB	5.7 GB	✅ Runs comfortably
z-image-Q3_K_S.gguf	Q3_K_S	4.4 GB	6.2 GB	✅ Runs comfortably
z-image-Q3_K_L.gguf	Q3_K_L	4.6 GB	6.5 GB	✅ Runs comfortably
z-image-Q4_0.gguf	Q4_0	4.6 GB	6.5 GB	✅ Runs comfortably
z-image-Q4_1.gguf	Q4_1	4.9 GB	6.8 GB	✅ Runs comfortably
z-image-BF16.gguf	GGUF	12.3 GB	15.0 GB	✅ Runs comfortably
model.safetensors	fp16	20.5 GB	24.0 GB	⚠️ Tight — reduce context

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

Specifications

Parameters: 6.2 B
Licence: apache-2.0
First seen on the Hub: 2026-01-27
Base model: Z-Image
Training datasets: undisclosed
Added to our catalog: 2026-08-21

Family

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

Base Z-Image Tongyi-MAI 44 K

Compare with any text-to-image model

Popular comparisons

Qwen-Image-2.1 10.2 M stable-diffusion-xl-base-1.0 4.6 M Krea-2 4.5 M Qwen-Image-2.1-Uncensored-Quantized 2.5 M Qwen-Image-2.1-Uncensored-GGUF 2.2 M Qwen-Image-2.1-Turbo-Quantized 90 K

Appears in

Commercial-use image models