> For the complete documentation index, see [llms.txt](https://scade.gitbook.io/scade-knowledge-base/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://scade.gitbook.io/scade-knowledge-base/build-a-flow/top-nodes-settings/image-generation.md).

# Image generation

Let's dive into the settings of image generation models like Focus and Juggernaut. Most of these models are based on Stable Diffusion. 

### Fooocus and other SD nodes

* Prompt: Describe what you want the image to look like, including content and style.
* Priority Queue: Set the order for processing requests.
* Negative Prompt: Specify what to avoid in the image to reduce mistakes like extra fingers.
* Style Selections: Choose the look or theme you want. Check GitHub for community examples.\
  <https://github.com/lllyasviel/Fooocus/discussions/2082>

<https://github.com/lllyasviel/Fooocus/discussions/143>

* Performance Selections: Adjust settings to balance speed and quality.
* Aspect Ratios: Choose the shape of the image.
* Image Number: Decide how many images you want to create.
* Image Seed: Use this for consistent results. The default is -1, but you can use specific numbers.
* Loras Custom URLs: Add external URLs for specific styles.
* Sharpness: Adjust how clear and detailed the image is.
* Guidance Scale: Balance between following the prompt and creative freedom.
* Refiner Switch: Choose whether to use a refining model to improve quality after generation.
* Uov Input Image: Select an image for transformations or guidance.
* Uov Upscale Value: Increase the image resolution.
* Inpaint Additional Prompt: Give extra details for filling in masked areas.
* Inpaint Input Image: The base image to be modified.
* Inpaint Input Mask: The mask defining which areas to change.
* Inpaint Strength: Control how much the inpainting follows the prompt.
* Outpaint Selections: Settings for extending the image beyond its original borders.
* Outpaint Distance: Specify how much to extend the image in each direction (left, top, right, bottom).
* Cn Prefix Fields: Upload images to mix and match, or use masks to add elements.

Similar concepts you can apply to other Stable Diffusion-like models as Juggernaut XL.

### DALL·E node

Let’s look at Dalle node fields:

* Prompt: Describe what you want the image to look like, including content and style.
* Model: Choose the model to create the image.
* N: Decide how many images to generate for comparison or selection.
* Quality: Choose the resolution or quality of the image.
* Response Format: Decide how the generated image will be returned.
* Size: Set the dimensions of the image.
* Style: Choose the artistic style or theme.
* Image: Provide an input image for transformation or as a base for new variations.
* Mask: Define which parts of the image to change.

Operation: Start the image creation process based on your specifications.