Samsa's image generation transforms your text descriptions into stunning visuals. Here's everything you need to know to start creating.
Write a prompt describing what you want to see
Choose your settings (aspect ratio, style, resolution)
Click Generate and your image appears in seconds
The prompt editor is your creative canvas. You can:
Write descriptions up to 5,000 characters
Use natural language - no special syntax needed
Reference your trained models using @mentions
Add negative prompts to exclude unwanted elements
Aspect Ratio
Select the shape of your image:
1:1 - Square, perfect for social media profiles
16:9 - Widescreen, great for presentations
9:16 - Vertical, ideal for stories and reels
4:3 - Standard landscape
3:4 - Standard portrait
4:5 - Instagram feed
21:9 - Ultra-wide cinematic
Style Models
Apply visual styles to your generations:
Use Samsa's base styles for instant aesthetics
Apply your custom-trained style models
Combine with person or object models
Resolution
Higher resolution = more detail (and more credits):
1K - 5 credits, great for drafts
2K - 10 credits, good for most uses
4K - 20 credits, best for final outputs
Base cost: 5 credits per image at 1K resolution
Be specific: "A golden retriever puppy" is good, "A golden retriever puppy playing in autumn leaves, soft afternoon light" is better
Describe the style: Include photography terms, art styles, or mood descriptors
Use references: Mention styles like "product photography" or "editorial portrait"
Iterate: Generate multiple versions and refine your prompt based on results
