Stable Diffusion Prompt Generator

Name

Matt Popovic

Updated on 7/24/2023

Welcome to the Randomized Prompt Generator for Stable Diffusion! Our prompts cover a range of genres and writing styles, so whether you're a fiction writer, poet, or non-fiction writer, you're sure to find a prompt that sparks your imagination.

So what are you waiting for? Give our Free Stable Diffusion Prompt Generator a try and fuel your creativity today!

How it works

We have collected the public data from Lexica.net (opens in a new tab) for the Top 100 keywords for Stable Diffusion prompts and classified them into four categories:

• Highly Detailed and Intricate. For example,

      'highly detailed',
      'sharp focus',
      'intricate',
      'artstation',
      'digital painting',

• Lighting and Atmosphere. Such as:

        'cinematic lighting',
        'dramatic lighting',

• Artist Style. e.g.

        'by dustin nguyen',
        'akihiko yoshida',
        'greg tocchini',
        'greg rutkowski',
        'cliff chiang',

• Image Specifications:

        '4k resolution',
        '8k',
        'Canon50',
        'Rendered by octane',
        'Close-up',

And we randomize the result by creating a Node.js component to mix the given input with these top keywords.

To-do: Plug into OpenAI API to use ChatGPT to streamline the result interactively. Right now, the input is pretty much irrelevant.

References

Read more about Stable Diffusion prompts generation:

• Stable Diffusion prompt: a definitive guide (opens in a new tab)
• Dataset extracted from Lexica.art at Hugging Face (opens in a new tab)
• Stable Diffusion Prompt Book - OpenArt (opens in a new tab)

Got suggestions?

If you have any suggestions, please drop them at our Discord (opens in a new tab).

Also, we are also contributing to Other AI-powered Open Source Projects: RATH (opens in a new tab).

RATH (opens in a new tab) brings the power of Open Source to the world of Data Analytics by Automating Data Exploration & Data Visualization down to a simple click. Feel free to check it out on GitHub (opens in a new tab)!