Z Image Engineer V6 DHQ needs about 4.2 GB of memory at Q4_K_M (a 2.5 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 9 quants, from 1.4 GB (IQ2_S) to 2.9 GB (Q5_K_M). Longer context needs more: at 32K tokens the KV cache alone adds about 4.8 GB.
Z Image Engineer V6 DHQ – a 4B Qwen3 fine‑tune for prompt enhancement and text encoding
Z Image Engineer V6 DHQ is a 4B Qwen3 model fine‑tuned for Z‑Image workflows. It functions as both a prompt writer and a text encoder. The DHQ quantization ensures each tensor has the precision needed for the Z‑Image DiT.
Good at
- Prompt enhancement for Z‑Image
- Text encoding for Z‑Image
- Local workflow without external telemetry
Limits
- No vision support
- Large context requires high VRAM
- License is apache‑2.0
Best for
- Local prompt rewriting
- Replacing Z‑Image text encoder
- Hybrid prompt‑to‑image pipelines
Who it’s for. Users of LM Studio, ComfyUI, or llama.cpp who need a local prompt enhancer and text encoder for Z‑Image. The model is suitable for artists and developers working with Z‑Image only.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ2_S | 1.4 GB | 3.1 GB | Fits | Get file → | |
| IQ2_M | 1.5 GB | 3.2 GB | Fits | Get file → | |
| Q2_K | 1.7 GB | 3.4 GB | Fits | Get file → | |
| IQ3_S | 1.9 GB | 3.6 GB | Fits | Get file → | |
| IQ3_M | 2.0 GB | 3.7 GB | Fits | Get file → | |
| Q3_K_M | 2.1 GB | 3.8 GB | Fits | Get file → | |
| IQ4_XS | 2.3 GB | 4.0 GB | Fits | Get file → | |
| Q4_K_M | 2.5 GB | 4.2 GB | Fits | Get file → | |
| Q5_K_M | 2.9 GB | 4.6 GB | Fits | Get file → |
For 8 GB VRAM, use IQ2_S (1.42 GB, 3.1 GB memory at 8k context) or IQ2_M (1.51 GB, 3.2 GB). If 12 GB VRAM is available, Q4_K_M (2.5 GB, 4.2 GB) offers a good balance of quality and memory. Larger models like Q5_K_M need 16 GB or more.
What GPU do you need for Z Image Engineer V6 DHQ?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is IQ4_XS (2.3 GB file, about 4.0 GB in use).
- 6 GB VRAM: the best quant that fits is Q5_K_M (2.9 GB file, about 4.6 GB in use).
Which Z Image Engineer V6 DHQ quant should you choose?
Q4_K_M is the usual starting point for Z Image Engineer V6 DHQ: it balances file size and quality well. Here it is a 2.5 GB file.
If memory is tight, the smallest file is IQ2_S at 1.4 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Z Image Engineer V6 DHQ
How much VRAM does Z Image Engineer V6 DHQ need?
About 4.2 GB at Q4_K_M with an 8K context window. That is the 2.5 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Z Image Engineer V6 DHQ on an 8 GB GPU?
Yes, with Q5_K_M: the 2.9 GB file needs about 4.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Z Image Engineer V6 DHQ GGUF?
It depends on the quant: IQ2_S is 1.4 GB, IQ2_M is 1.5 GB, Q2_K is 1.7 GB, IQ3_S is 1.9 GB, IQ3_M is 2.0 GB, Q3_K_M is 2.1 GB and 3 more.
What context length does Z Image Engineer V6 DHQ support?
The GGUF header lists a trained context length of 40,960 tokens (40K). The KV cache grows with the context you use: about 1.2 GB at 8K and 4.8 GB at 32K.
Where can I download Z Image Engineer V6 DHQ GGUF?
The files are on Hugging Face in the pottokao/Z-Image-Engineer-V6-DHQ-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Z Image Engineer V6 DHQ use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Z Image Engineer V6 DHQ on 8 GB VRAM?
Yes, the IQ2_S or IQ2_M quantizations fit 8 GB VRAM with about 3.1–3.2 GB memory usage at 8k context.
Does the model support image generation directly?
No, it is a text encoder and prompt writer; it does not generate images itself.
Is the model licensed for commercial use?
It is licensed under apache‑2.0, which allows commercial use with attribution.