Stable Diffusion Prompt Commands

Introduction to Stable Diffusion Prompt Commands

Stable Diffusion is a powerful AI image generation tool that creates images based on text prompts. Writing effective prompts is key to generating high-quality images that match your vision. This guide will provide an overview of useful prompt commands and examples to help you get the most out of Stable Diffusion.

Basic Prompt Structure

A basic Stable Diffusion prompt has a simple structure:

"A [style] [subject] [modifiers]" 

For example:

"A oil painting of a cat wearing a top hat and monocle"

Let’s break down the key components:

  • Style: The desired artistic style, like “oil painting” or “digital art.”
  • Subject: The main focus of the image, like “cat” or “dragon.”
  • Modifiers: Extra descriptive details about the subject, scene, framing, etc.

Modifiers are important for adding specificity to get better results. Now let’s look at some advanced prompt engineering techniques.

Prompt Commands

Stable Diffusion supports various prompt commands to control aspects of the generated image:

--ar [aspect ratio]
--upbeta 
-upbeta
--creative
--testp 

Here’s what they do:

  • --ar: Sets aspect ratio, like --ar 16:9.
  • --upbeta: Enhances image coherence.
  • -upbeta: Reduces image coherence.
  • --creative: Increases image variety.
  • --testp: Evaluates prompt first.

For example:

"--ar 16:9 --upbeta --creative A futuristic cityscape with flying cars"

This generates a wide 16:9 image with increased coherence and variety.

Keyword Modifiers

You can also modify keywords in the prompt to influence the AI:

(keyword:weight)
[keyword] 
  • (keyword:weight): Increases attention on that keyword, higher weight = more focus.
  • [keyword]: Decreases attention on that keyword.

For example:

(city:2.0) (flying cars:1.5) [trees] A futuristic cityscape with flying cars

This puts extra focus on the city and cars, and reduces emphasis on trees.

Helpful Prompt Examples

Here are some prompt examples to help you generate better images with Stable Diffusion:

Styles

Specify an artistic style:

"A retro futuristic digital painting of a robot butler" 

"An ukiyo-e style image of a samurai holding a laser katana"

Framing

Set the framing and composition:

"A close-up portrait photo of an astronaut in a spacesuit"

"A full body photo of a wizard casting a spell, shot from a low angle perspective" 

Lighting

Add lighting details:

"A 3D render of a dragon lit dramatically with a rim light from the side"

"A photo at sunset of a lonely house on a hillside with warm evening light"

Subject Details

Include extra subject descriptors:

"An oil painting of a powerful knight wearing intricate armor"

"A cute capybara wearing sunglasses and a colorful summer shirt sitting in the grass"

Conclusion

Mastering prompts for Stable Diffusion takes experimentation, but this guide should provide a strong starting point. Focus on adding stylistic details, framing instructions, lighting cues, and descriptive modifiers to get better results. Check out Stable Diffusion Prompt Hero and Stable Diffusion Art for more examples and inspiration!