Image to Image lets you transform an existing photo using a text prompt. Instead of generating from scratch, you guide edits on top of your original image.
What is Image to Image?
It keeps key elements intact—faces, composition, and structure—while changing the parts you describe, like background, lighting, colors, or style.
Perfect for product shots, portraits, concept art, and style transfer when you need control and consistency.
Transform an Image in 4 Steps
Upload your image, describe the changes, and let the AI refine your photo while preserving what matters.
- 1
Step 1: Upload Your Image
Choose one or multiple images to use as references.
- 2
Step 2: Describe the Edit
Write clear instructions such as “keep the subject, change the background to a snowy forest.”
- 3
Step 3: Set Output Options
Pick aspect ratio and the number of results you want.
- 4
Step 4: Generate & Download
Review the edits, then download your favorites.
Why Use Image to Image?
Get precise control over edits without re-creating your image from scratch.
Targeted Edits
Change specific parts like background, colors, or mood while keeping the subject intact.
Structure Preservation
Maintain composition, pose, and layout for consistent results across variations.
Multi-Image Reference
Use multiple images to transfer style or match a desired look.
Fast Iteration
Create several variations quickly to explore ideas and finalize the best result.
Transform Your First Image
Edit photos with simple instructions and get variations in seconds.
Image to Image FAQ
Answers to common questions about editing images with AI.
What can I change with image to image?
You can alter backgrounds, colors, lighting, styles, and many details while keeping the core subject.
How do I keep certain elements unchanged?
Be explicit in your prompt: for example, “keep the face, outfit, and pose; change only the background.”
Can I use multiple images?
Yes, you can upload multiple reference images to guide the style or composition.
What image formats are supported?
Common formats like JPEG, PNG, and WEBP are supported.
How many outputs can I generate?
Choose 1, 2, or 4 outputs per run.
Is the result commercial-ready?
Paid users can use the results commercially.
How long does it take?
Most edits finish in seconds to a minute depending on complexity.
What if the result isn’t right?
Refine your prompt or upload a clearer reference and try again.