---
base_model:
- Tongyi-MAI/Z-Image-Turbo
---
GGUF quantized version of the Z-Image text encoder (Qwen 3 4b)



Recommended:
- Qwen_3_4b-imatrix-IQ4_XS.gguf - default pick
- Qwen_3_4b-imatrix-Q3_K_M.gguf - if you really need that extra 200mb
- Qwen_3_4b-Q8_0.gguf - near identical results to FP16

![comparison](https://cdn-uploads.huggingface.co/production/uploads/66edc6ec77590e90ba7c109c/G0Kf_tsEUtiGMIb1Ksyh5.png)