Stable Diffusion XL (SDXL)
Stable Diffusion XL é a versatile, open-source model known for its artistic flexibility and strong community support.
Visão geral
| Propriedade | Valor |
|---|---|
| Provider | Stability AI |
| Custo | 40 credits |
| Modalidade | Image |
| Imagens de referência | ❌ None |
| Prompt obrigatório | Yes |
Para que é melhor
- Artistic styles — Wide range of artistic interpretations
- Creative experimentation — Flexible and forgiving
- Concept art — Illustrations and artistic renders
- Style variety — From photorealistic to abstract
- Community resources — Extensive prompt guides available
Inputs
Prompt (Required)
Describe the image you want to generate.
Cor da conexão: 🟡 Yellow
Configuração
Width / Height
Type: Select
Options: 512, 768, 1024
Default: 1024
Negative Prompt
Type: Textarea
Describe what you don't want in the image.
Inference Steps
Type: Slider
Range: 1 - 100
Default: 50
More steps = better quality but slower.
Guidance Scale
Type: Slider
Range: 1 - 20
Default: 7.5
How closely to follow the prompt:
- Low (1-5): More creative freedom
- Medium (6-10): Balanced
- High (11-20): Strict prompt adherence
Output
Type: Image
Cor da conexão: 🟢 Green
Casos de uso
Digital Art
Fantasy landscape with floating islands, magical crystals,
ethereal lighting, digital painting style, highly detailed.
Character Illustration
Anime-style warrior character, detailed armor, dynamic pose,
vibrant colors, professional illustration.
Concept Art
Sci-fi spaceship interior, control room, holographic displays,
moody lighting, concept art style.
Abstract Art
Abstract fluid art, vibrant colors, flowing shapes,
marble texture, high contrast.
Negative Prompt Tips
Use negative prompts to avoid common issues:
blurry, low quality, distorted, deformed, ugly,
bad anatomy, extra limbs, watermark, text, signature
For portraits:
bad hands, extra fingers, missing fingers,
crossed eyes, asymmetric face
Dicas para melhores resultados
- Use negative prompts — Essential for SDXL
- Specify art style — "digital painting", "oil painting", "anime"
- Include quality keywords — "highly detailed", "professional"
- Adjust guidance scale — Higher for specific results
- Experiment com steps — 50 é a good starting point
Comparação com outros modelos
| Feature | SDXL | Flux 2 Flex | Qwen Image |
|---|---|---|---|
| Cost | 40 credits | 20 credits | 20 credits |
| Style Range | ⭐⭐⭐⭐⭐ | ⭐⭐⭐⭐ | ⭐⭐⭐ |
| Photorealism | ⭐⭐⭐ | ⭐⭐⭐⭐ | ⭐⭐⭐ |
| Text Rendering | ⭐⭐ | ⭐⭐⭐ | ⭐⭐⭐⭐ |
| References | ❌ | 10 | 1 |
When to Choose SDXL
Choose SDXL when:
- You want artistic/creative styles
- You're familiar com SD prompting
- You want community-tested prompts
- Artistic interpretation é desired
Consider alternatives when:
- You need photorealism (Flux models)
- You need text in images (Qwen, Nano Banana)
- You need imagens de referência
Modelos relacionados
- Flux 2 Flex — More photorealistic
- Qwen Image — Better text, similar price
- Flux Kontext Pro — Style transformations