ArtBot

Note this is WIP

Using

(Recommended) Sign In first

Create a new image

• Prompt: Enter descriptive text outlining the things that you would like in your image. For example TODO
• Negative Prompt: Enter descriptive text outlining the things that you would not like in your image. A good starting point might be

  Worst quality, low quality, deformed, blurry, deformed eyes, deformed face, monochrome, missing fingers, extra didgit, fewer didgits, extra body parts, bad anatomy, censored, collage, logo, border 
  

• Sampler:
• Number of Images
• Image Models: Models can be found on https://civitai.com/models (note stablehorde supports a subset of these). Artbot also allows you to search these
• Image Orientation: Choose between some preset options or use custom. Note that stable diffusion is trained on 512x512 so this will likely give you best results
• Steps
• Guidance
• Seed
• Clip Skip
• Slect LoRAs (max 5)
• Textual Inversions
• Post processing
• Upscalers
• Additional Options:
  • Karras
  • Hires fix
  • Use slow workers
  • Tiling
  • Allow NSFW Generations: Enable NSFW (note fewer generators support this)
  • Use Prompt Replacement Filter