| --- |
| license: apache-2.0 |
| language: |
| - en |
| pipeline_tag: text-to-image |
| base_model: Tongyi-MAI/Z-Image |
| tags: |
| - gguf |
| - safetensors |
| - text-to-image |
| - rundiffusion |
| - z-image |
| --- |
| |
| <h1 align="center">Juggernaut Z<br><sub><sup>A polished cinematic fine-tune of Z-Image Base from RunDiffusion</sup></sub></h1> |
|
|
| <div align="center"> |
|
|
| [](https://www.rundiffusion.com/juggernaut-z) |
| [](https://huggingface.co/Tongyi-MAI/Z-Image) |
| [](https://www.rundiffusion.com/juggernaut-z-prompt-guide) |
|
|
| </div> |
|
|
| Juggernaut Z is a fine-tuned image model built on **Z-Image Base**, created through the work of **Team Juggernaut** with fine-tuning by **KandooAI**. On RunDiffusion, it is positioned as a stronger choice for creators who want more polished image output, better lighting quality, stronger camera focus, and more detailed skin textures. |
|
|
| This repository is intended to host the RunDiffusion release artifacts for Juggernaut Z, including full-precision weights and GGUF quantizations. |
|
|
|  |
|
|
| ## Overview |
|
|
| Juggernaut Z is tuned for a more presentation-ready look out of the box. Relative to Z-Image Base, the emphasis is on: |
|
|
| - Stronger lighting and clearer atmosphere |
| - More refined focus and camera feel |
| - More polished portrait rendering |
| - Improved skin texture detail |
| - Better out-of-the-box presentation for editorial, concept, and cinematic work |
|
|
| ## Website Comparisons |
|
|
| The comparison sets below are taken from the RunDiffusion Juggernaut Z announcement page. |
|
|
| **Comparison label:** Left: **Juggernaut Z** · Right: **Z Image Base** |
|
|
| ### Better Lighting |
|
|
| More dramatic, cinematic lighting quality out of the box. |
|
|
|  |
|  |
|  |
|  |
|  |
|  |
|
|
| ### Improved Textures |
|
|
| Juggernaut Z produces cleaner and more natural-looking skin textures, especially in close-up portraits. |
|
|
|  |
|  |
|  |
|  |
|
|
| ### Anatomy |
|
|
| We have worked to improve overall image integrity, resulting in cleaner anatomy and more consistent structural detail across a wide variety of subjects. |
|
|
|  |
|  |
|  |
|  |
|
|
| ### Composition |
|
|
| Subject and object placement within scenes improved in version one, with continued improvements planned for version two. |
|
|
|  |
|  |
|  |
|
|
| ### Diversity |
|
|
| The model is tuned toward a more balanced result across ethnic backgrounds, with better representation out of the box. |
|
|
|  |
|  |
|  |
|  |
|
|
| ### Architecture |
|
|
| Architectural subjects benefit from cleaner structural lines and more coherent material rendering. |
|
|
|  |
|  |
|
|
| ## Recommended Settings |
|
|
| Based on the RunDiffusion launch guidance: |
|
|
| - Recommended default: `CFG 6`, `35 steps` |
| - Good CFG range: `6 to 9` |
| - Good steps range: `25 to 45` |
|
|
| ## Good Fit For |
|
|
| - Portraits with cleaner facial detail and stronger focus |
| - Cinematic scenes with stronger lighting and clearer atmosphere |
| - Concept development and visual exploration |
| - Editorial and fashion work that benefits from a polished finish |
|
|
| ## Files In This Repo |
|
|
| Current release artifacts: |
|
|
| - `Juggernaut_Z_V1_by_RunDiffusion.safetensors` |
| - `Juggernaut_Z_V1_FP8_e4m3fn.safetensors` |
| - `Juggernaut_Z_V1_by_RunDiffusion_q4_k_s-001.gguf` |
| - `Juggernaut_Z_V1_by_RunDiffusion_q4_k_m-002.gguf` |
| - `Juggernaut_Z_V1_by_RunDiffusion_q5_k_s-005.gguf` |
| - `Juggernaut_Z_V1_by_RunDiffusion_q5_k_m-003.gguf` |
| - `Juggernaut_Z_V1_by_RunDiffusion_q6_k-004.gguf` |
| - `Juggernaut_Z_V1_by_RunDiffusion_q8_0.gguf` |
|
|
| ## Notes |
|
|
| - The base model is [Tongyi-MAI/Z-Image](https://huggingface.co/Tongyi-MAI/Z-Image). |
| - This card uses image assets from the RunDiffusion Juggernaut Z announcement page. |
| - If you are using the GGUF files, use a GGUF-compatible runtime or workflow. |
| - If you are using the safetensors releases, load them with the workflow that matches your local inference stack. |
|
|
| ## Links |
|
|
| - Announcement: [RunDiffusion Juggernaut Z](https://www.rundiffusion.com/juggernaut-z) |
| - Prompt guide: [Juggernaut Z Prompt Guide](https://www.rundiffusion.com/juggernaut-z-prompt-guide) |
| - Base model: [Tongyi-MAI/Z-Image](https://huggingface.co/Tongyi-MAI/Z-Image) |
|
|
| ## Attribution |
|
|
| Juggernaut Z is based on Z-Image Base. Credit for the upstream base model belongs to the Z-Image team. Credit for this fine-tuned release is attributed to Team Juggernaut, with fine-tuning by KandooAI, as described on the RunDiffusion announcement page. |
|
|
