High-quality, detailed outputs. Best for final renders.
Input Image
Drop image or browse
Mask Image
How masks work:
⬜ White pixels → New AI generation
⬛ Black pixels → Preserved from original
⬜ White pixels → New AI generation
⬛ Black pixels → Preserved from original
Drop mask or browse
Width
1024px
Height
1024px
Number of Images
1
Strength
0.75
Controls how much the generated image differs from the input. Higher = more variation from input.
Fix the seed to reproduce identical outputs.
Inference Steps
28
Higher steps = better quality but slower. Lower steps = faster but lower quality.
Guidance Scale
3.5
Controls adherence to prompt. Higher = more faithful to prompt description.
Ctrl + ↵ to generate
Ready to create
Configure your settings, enter a prompt, and hit Generate to bring your vision to life.
Generating…
Uploading assets and queuing prediction
Upload media assets
Queue generation request
Poll for results