Image-to-Image Node

Transform existing images using AI while maintaining structure or applying new styles.

Overview

The Image-to-Image node modifies existing images based on text prompts. Use it to create variations, apply styles, or make specific changes while keeping the original composition.

Inputs

Source Image (required)

Connect from another node or upload
Any image format (JPG, PNG, WebP)

Text Prompt (required)

Describes desired changes
Can connect from Prompt Node

Outputs

Transformed Image

Modified version of input image
Saved to Media Library

Configuration

Transformation Strength

0-100% - How much to change the image

Low (0-30%): Subtle changes, keep original closely
Medium (30-70%): Balanced transformation
High (70-100%): Major changes, loose interpretation

Model Selection

FLUX Dev Image-to-Image
Leonardo AI variants
Stable Diffusion models

Use Cases

Style Transfer

Input: Photo of building Prompt: "Same building in watercolor painting style" Result: Artistic version maintaining structure

Variations

Input: Character design Prompt: "Same character in different pose" Result: New variation of character

Refinement

Input: Draft image Prompt: "More detailed, professional quality" Result: Enhanced version

Background Change

Input: Product photo Prompt: "Same product on wooden table outdoors" Result: New background, same subject

Workflow Integration

Input from:

Text-to-Image
Asset Node
Explore Node

Output to:

Image-to-Video
Multi-Angle Generator
Another Image-to-Image (chain)

Tips

Start with 50% strength and adjust
Be specific about what to keep vs change
Use "same [subject]" in prompt to maintain subject
Chain multiple Image-to-Image nodes for gradual changes