Home Artists Posts Import Register
Patreon importer is back online! Tell your friends ✅

Downloads

Content

Hey everyone! I've created a cool new resource for Stable Diffusion users - a style.csv file packed with 100 different art styles to use in your image generation.

This file contains prompts for styles ranging from Realism to Pop Art, Traditional Japanese Art to Graffiti. Some examples:

  • Retro Video Game: pixelated, 8-bit, arcade style
  • Cubist: Picasso-esque, abstract, geometric
  • Ukiyo-e: traditional Japanese woodblock print
  • Low Poly: simple geometric 3D

To use the file, simply download style.csv from the Patreon post and put it in your root /stable-diffusion-webui folder. Then in the WebUI, you'll be able to select from any of these 100 styles in the dropdown menu.

Having these organized prompts makes it super easy to get a huge variety of art styles applied to your image generation. Whether you want a Ukiyo-e take on your prompt or an 80s pixel art version, you can get amazing stylized results.

Let me know if you have any requests for styles you'd like me to add in future updates!

And as always, supporting me on Patreon allows me to keep creating helpful resources like this for the AI art community. Thank you for your support - now go make some awesome art

Files

Comments

Rai T00n

My AI overlord, this is what every ordinary people want but no experts were making. I was using portable comfyUI only because of the complex install process of 1111. Now I can finally try 1111. Thank you!

Anonymous

Do these styles work on every checkpoint? Or just the default one?

Anonymous

My UI doesn't have the apply button. It only has the first two (downward arrow, and wastebasket). How do I see the other options?

Rubén Darío Mena

I don't know how I spend so much time without using this! Thank you!

antonde slovenia

hello and thanks for your work. The installation process went smoothly. unfortunetely my images dont look even close to realistic or photo-like. no matter what promts i use or what styles i apply. the images look like from under-trained models. what can be a sollution? they are too simple https://anton-deineko.nimbusweb.me/share/10419375/7o6twv8k3a2hyes2ze2k

Aitrepreneur

it's because you need to learn how to prompt for it, this is not midjourney, there is a huge learning curve. For example I can see that you are using an sdxl model and yet your resolution is 512:512, you need to use a resolution superior to 1024:1024 with sdxl models. You can also dl other stable diffusion models trained by the community that you can find on civitai.com and copy the generation data