Stable Diffusion Prompt Generator
Build professional SDXL / SD 1.5 prompts with guided fields, negative prompt presets, and 5 style variations.
Quick Style Presets
Prompt Builder
Realistic to SDXL (1024px base) - Anime / Art to SD 1.5 (512px base)
Generation Parameters
Output: 1024Ã-1024px
Recommended: 6-9 for most use cases
Recommended: 25-35
How to Use
- Choose a Model - Realistic for SDXL photorealism, Anime for anime style, Art for general creative work.
- Apply a Style Preset to auto-fill style, lighting, and camera settings as a starting point.
- Enter your Subject (required) and Scene to describe what you want to generate.
- Fine-tune Art Style, Lighting, Composition, Camera, Color, and Mood using the dropdowns.
- Add an Artist Reference for a specific aesthetic (e.g., "Greg Rutkowski", "Artgerm").
- Select a Negative Preset to avoid common issues (blurry faces, bad anatomy, watermarks).
- Adjust CFG Scale, Steps, Aspect Ratio, and Sampler for your target model.
- Click Generate Prompts - copy the main prompt or any of the 5 style variations.
FAQs
Pro Tips
Be specific with subjects. "A 25-year-old woman with shoulder-length brown hair, green eyes, wearing a red dress" beats "a woman" every time.
Lighting transforms images. Adding "golden hour lighting", "dramatic rim light", or "volumetric lighting" is one of the highest-impact changes you can make.
Use 2-3 quality tags max. "masterpiece, best quality, highly detailed" is enough. Stacking 10 quality tags causes oversaturation.
Iterate with the same seed. Once you find a good composition, lock the seed and tweak the prompt to refine details without losing the layout.
Batch generate. Run 4-8 images at once to get variety, then pick the best result and use its seed for further refinement.