Table of Contents
Custom AI Model Training: Complete Guide
Learn how to train and use your own custom AI models for generating unique, personalized images that match your specific style, product, or person.
Understanding Custom Models
Transform your creative workflow with custom AI models. Whether you're showcasing products, replicating artistic styles, or generating consistent character images, custom models provide the foundation for precise, personalized AI image generation.
Available Model Types
Product Models
Perfect for e-commerce, product photography, and marketing materials
Style Models
Capture unique artistic styles, textures, or design aesthetics
Person Models
Train the AI to generate images of specific individuals
Training Process Guide
1. Model Setup
Essential Configuration
Name Your Model
- Choose a descriptive, memorable name
- This becomes your trigger word for generation
- Keep it simple and relevant
Select Model Type
- Product: Physical items and merchandise
- Style: Artistic and design patterns
- Person: Individual character training
2. Image Preparation
Technical Requirements
- Minimum: 5 high-quality images
- Recommended: 10-20 images for best results
- Maximum file size: 10MB per image
- Formats: PNG or JPG
Optimization Guidelines
Product Model Best Practices
- Capture multiple angles (front, side, back)
- Maintain consistent lighting conditions
- Use clean, neutral backgrounds
- Include both full product and detail shots
- Show the product in its intended context
Style Model Best Practices
- Include diverse style examples
- Maintain consistent artistic elements
- Provide both simple and complex compositions
- Show style applied to different subjects
- Include signature elements of the style
Person Model Best Practices
- Capture various facial expressions
- Include different camera angles
- Mix head shots and full-body poses
- Use natural lighting and poses
- Avoid heavy makeup or unusual accessories
Using Your Models
Training Process
What to Expect
Training Duration
The training process typically takes 5-10 minutes. During this time:
- Your images are processed and analyzed
- The AI learns your subject's unique characteristics
- The model is optimized for generation
Training Status
Your model will display one of these statuses:
- Training: Model is currently being trained
- Ready-to-use: Training completed successfully
- Failed: Training encountered an error
After Training
- Look for the "Ready-to-use" status in your dashboard
- Click the "Use" button to start generating images
- Generate variations using different prompts with your model
Model Management
Custom Models
Model Cards
Each model in your Custom Models page displays:
- Preview image of your trained subject
- Model name
- Creation date for easy reference
- Current status indicator
Quick Actions
For ready-to-use models, you can:
- Click "Use" to immediately start generating images
- Access the image generator with your model pre-selected
- View model details and training information
Generating Images
Generation Process
Model Selection
- Choose your trained model from the dropdown menu
- Models are labeled with their type (Style, Person, or Product)
- Each model maintains its specific characteristics during generation
Prompt Writing
Different model types require different prompt approaches:
Style Models
Focus on describing the scene or subject you want to create:
- "A majestic mountain landscape with snow-capped peaks"
- "Modern cityscape at golden hour"
- "A bouquet of wildflowers in a vintage vase"
Person Models
Describe poses, expressions, and settings:
- "Professional headshot with a warm smile, office background"
- "Casual outdoor portrait in a park, candid pose"
- "Close-up shot with dramatic side lighting"
Product Models
Focus on presentation and context:
- "Clean product shot on white background, front view"
- "Lifestyle photo in modern living room setting"
- "Close-up detail shot highlighting texture"
Image Settings
- Aspect Ratio: Choose from multiple options (1:1, 16:9, etc.)
- Number of Images: Generate up to 4 variations at once
- Each aspect ratio is optimized for specific use cases
Working with Generated Images
Available Actions
Image Operations
- Download: Save images locally in high quality
- Visual Muse: Get inspiration for variations
- Reimagine: Create new versions of the image
- Shuffle: Generate alternative versions
- Upscale: Enhance image resolution
- Expand: Modify image dimensions
Preview Features
- View full-size images in lightbox mode
- Navigate between multiple generated variations
- Quick thumbnail selection for multiple images
- Retry loading for failed images
Tips for Best Results
- Be specific in your prompts to get more accurate results
- Experiment with different aspect ratios for optimal composition
- Generate multiple images to get more variations
- Use the preview features to compare and select the best results
- Try different prompt structures if initial results aren't satisfactory
Plan Limitations
Pro
4 model trainings per month
Premium
10 model trainings per month
Business
20 model trainings per month
Troubleshooting Guide
Common Issues and Solutions
Training Failures
- Check image quality and quantity
- Ensure consistent subject matter
- Verify file formats and sizes
Related Content
AI Image Creator: Your Complete Guide
Learn how to use CreativePixel to create custom images. This comprehensive guide covers all aspects of the AI image creation process.
Read More →Reimagine Image Tool: Your Complete Guide
Learn how to use our AI-powered Image Reimagination Tool to transform your images based on your ideas. Perfect for designers, artists, and creative enthusiasts.
Read More →Image Upscale: Enhance Your Images with AI
Learn how to use our AI-powered Image Upscale tool to increase the resolution and quality of your images. Perfect for professionals and enthusiasts alike.
Read More →