Stable Diffusion Prompts & Settings
Stable Diffusion rewards precise, comma-separated tags plus a strong negative prompt more than natural sentences. These formulas cover the most common styles and the settings that fix typical flaws.
Photorealistic
photo of [SUBJECT], 85mm lens, natural window lighting, shallow depth of field, ultra detailed, 8k, professional photographyportrait of [SUBJECT], studio lighting, Rembrandt lighting setup, sharp focus on eyes, skin texture detail, DSLR photo
Illustration & Art Styles
[SUBJECT], digital painting, trending on artstation, dramatic lighting, highly detailed, concept art[SUBJECT], watercolor painting, soft edges, paper texture, muted color palette, loose brushstrokes[SUBJECT], anime style, cel shaded, vibrant colors, clean lineart, studio quality
Negative Prompts (paste alongside any prompt above)
blurry, low quality, distorted, deformed hands, extra fingers, watermark, text, signature, cropped, out of frameoversaturated, low contrast, jpeg artifacts, duplicate, mutated, bad anatomy, poorly drawn face
Tips for Better Results
- Always pair a positive prompt with a negative prompt — it fixes the majority of common flaws (bad hands, watermarks, blur).
- Order matters: put the most important subject/style words first — later tokens carry less weight.
- Use a CFG scale around 7-9 for balanced prompt adherence; higher values overcook the image, lower values ignore your prompt.
- Pick a checkpoint/model suited to your style (photoreal vs anime vs illustration) — the same prompt varies hugely across models.
New to Stable Diffusion? Start with the step-by-step guide.
◆ HOW TO USE STABLE DIFFUSION