Juggernaut XI v11

Juggernaut XI v11

A text-to-image model from RunDiffusion, part of the Juggernaut SDXL line, tuned for photo-style pictures with better hands, faces and prompt following. For image makers using ComfyUI, Forge and similar tools.

Made by RunDiffusion on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: Juggernaut XI v11 is a text-to-image model from RunDiffusion, based on Stable Diffusion XL. It was trained with GPT-4 Vision captions, and this round adds more dataset cleanup and sharper shot-type labels. The card says it improves hands, faces, eyes, composition and prompt following, and that it can render short legible text.

Who it suits: Image makers who want photo-style results from an SDXL model and work in a common Stable Diffusion interface. The license limits commercial use, so it suits personal and non-commercial work.

Industry | Who uses it | How they use it | Example
Digital art | Illustrators and hobbyists | Generate photo-style pictures from a written prompt | A portrait or full-body shot from a short description
Concept design | Designers drafting visuals | Sketch scenes before real production | A mood image for a location or a lighting style
Research | People comparing SDXL fine-tunes | Test prompt following and anatomy | Run the same prompt in natural language and in tags

You can skip it if you need a model licensed for commercial work or you want a speedy one, since the card points to a separate Lightning variant for that.

Where to use it:

• ComfyUI, Forge, Automatic1111, Fooocus, InvokeAI and SwarmUI, which RunDiffusion says have it pre-loaded
• RunDiffusion's hosted service in a browser, with a free trial
• Your own computer on Windows, macOS or Linux with Python, using the diffusers library

Use case | Who | What they get back
Photo-style pictures | Image makers | A rendered image from your prompt
Short text in an image | Designers | Legible signage or labels in the picture
Compare prompting styles | Prompt writers | Results from natural language or tag-style prompts

How to set it up: There are two routes. The hosted route needs only a RunDiffusion account. The local route needs access to a gated Hugging Face repository and enough GPU memory for an SDXL model.

• Python and pip for the local route
• A Hugging Face account that has been granted access
• Disk space and a GPU for SDXL-sized files

1. Request access on the model page.
2. Install the Hugging Face hub package.
3. Download the repository.
4. Point your image tool at the downloaded files.

pip install -U huggingface_hub
hf download RunDiffusion/Juggernaut-XI-v11

How to use it: Write your prompt in either of two styles, plain sentences or comma-separated tags. Then set the generation values the card recommends. The VAE is already baked in, so you need no separate one.

Setting | Value
Resolution | 832 by 1216 for portrait, 1216 by 832 for wide pictures
Sampler | DPM++ 2M Karras
Steps | 30 to 40
CFG scale | 3 to 7, lower is more realistic
VAE | Already included

Example: In your image tool you type a prompt such as "a lighthouse at dusk, waves hitting the rocks, shot like a film photograph". You choose the wide size, the sampler and about 30 steps. An image of that scene comes back at the size you chose.

For newcomers and experts: If you are new, the easiest start is the hosted route in a browser, where nothing needs installing. Type a short description of a scene and look at what the model draws before you change any settings.

For experts, this is an SDXL fine-tune trained from the ground up on GPT-4 Vision captions, with refined shot-type classification and an aesthetic target. The card pairs it with a RunDiffusion Photo detail pass. The trade-off is licensing, since the terms limit commercial use and edits, and the gated download adds a step.

Written prompt
SDXL-based Juggernaut XI
Baked-in VAE
Image at your chosen size

Good to know:

• Access: the Hugging Face copy is gated, and the hosted route skips the gate
• Speed: a separate Lightning variant is aimed at 5 to 7 step runs
• License: CC BY-NC-ND 4.0, so read its terms before any commercial use
• Cost to run: the hosted route has a free trial, and the card names no other fee
• Data it reaches: downloads from Hugging Face, or your prompts go to RunDiffusion if you use the hosted route
• Not for: commercial work or sharing altered versions, as the license name indicates

On Hugging Face, 13 November 2026

• Task: text to image
• Library: diffusers
• Downloads in the last 30 days: 28,120
• Likes: 169
• License: cc-by-nc-nd-4.0
• Languages: en
• Based on: stabilityai/stable-diffusion-xl-base-1.0
• Fine-tunes: 1
• Tags: diffusers, stable-diffusion, stable-diffusion-xl, sdxl, text-to-image, photorealistic, photography, cinematic, portrait, juggernaut, rundiffusion, kandooai

Links

• Model: https://huggingface.co/RunDiffusion/Juggernaut-XI-v11
• Files: https://huggingface.co/RunDiffusion/Juggernaut-XI-v11/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download RunDiffusion/Juggernaut-XI-v11

Then follow the usage in its model card. Nothing in this kit runs by itself.