Add your source
Upload the source image that gives the tool the subject, composition, or visual reference to work from.
Use a source image and an edit prompt to create focused edited variants for a project, campaign, concept, or visual direction.
Image to Image is a creation mode that uses an existing visual plus a written prompt to produce edited variants based on your requested direction.

Use the guided online flow to transform image with a prompt, then compare the generated variants against your intended direction.
Upload the source image that gives the tool the subject, composition, or visual reference to work from.
Describe the edit, including style, subject, setting, and anything that should remain recognizable.
Select the destination format so the created variants better match your intended placement.
The image to image AI flow supports practical controls for directing how the source visual changes while keeping your project goal in view.

Write a detailed prompt that explains what should change, what should stay recognizable, and what the final visual should communicate.

Set a clear creative direction for style transfer, mood changes, background replacement, or other guided edits.

Describe the subject and setting so the generated variants have a stronger connection to your intended scene.
Create or use a result for practical visual tasks where an existing asset needs a guided change instead of a full restart.

Apply a new look, mood, or art direction to an existing visual while using the source as the starting point.

Guide a localized edit by describing what should appear in a selected area or changed part of the scene.

Replace or redirect the setting behind a subject to better fit a campaign, product page, or social post.
Before selecting a final variant, review the output against the brief, key details, and destination format.

Compare the variant against the original goal, campaign need, or creative brief before moving forward.
Inspect important details such as subject shape, background elements, visual consistency, and any areas changed by the prompt.
Check whether the destination crop and format match where the final asset will be used.
Generated results can vary, and important details may need careful checking before you use a variant in a final layout.
Two runs with similar prompts may not produce the same result, especially for complex style, setting, or subject changes.
Small text, hands, logos, faces, or product-specific details may need inspection before a variant is ready.
Answers to common questions about Image to Image inputs, prompts, outputs, and practical use.
Upload a source visual, describe the edit you want, and start creating project-ready variants.