SDXL 1.0

image

by Stability AI · Updated June 1, 2025

Stable Diffusion XL 1.0 is the industry-standard open-source image generator. With a massive ecosystem of community models, LoRAs, and tools, SDXL offers unparalleled customization and control over the generation process.

Best For

Community ecosystemCustom modelsLoRA fine-tuningControlNet workflowsLocal generation

Prompting Tips

1Use weighted prompts: (important element:1.3) to emphasize
2Negative prompts are essential: specify what to avoid
3Combine with LoRAs for specific styles and characters
4Use CFG scale 5-9 for best results (7 is a good default)
5Higher step counts (30-50) improve quality
6Use CLIP skip 1-2 for most models

Syntax & Constraints

Weighted prompt syntax: (element:weight). Negative prompt supported. Parameters: CFG scale (1-30), steps (20-50), sampler, seed, size (1024x1024 native).

Example Prompts

◆ Logo Design

Geometric mountain logo

minimalist geometric logo, clean vector design, abstract mountain peak symbol, blue and white color scheme, professional branding, centered composition, flat design, white background Negative: text, letters, words, blurry, low quality, 3d render

geometricminimalistmountain

👤 Portrait Photography

Professional studio portrait

(masterpiece, best quality:1.2), professional portrait photograph, young woman, natural lighting, soft bokeh background, (sharp focus on eyes:1.1), neutral color grading, 85mm lens, f/1.8, studio quality Negative: (worst quality, low quality:1.4), blurry, deformed, disfigured, bad anatomy, bad hands

portraitstudioprofessional

🏔 Landscape & Scenery

Patagonia epic landscape

(masterpiece, best quality:1.2), epic landscape photography, Patagonian mountains, dramatic storm clouds parting to reveal sunlight, turquoise glacial lake in foreground, (golden hour light:1.1), ultra-detailed, wide angle lens Negative: (worst quality, low quality:1.4), blurry, oversaturated, cartoon, painting

mountainsepicstorm