A familiar image.
A new direction.

What is image-to-image AI?

Image-to-image generation uses source images and a written instruction to produce a new still image. The sources provide visual context while the prompt describes the intended changes. It is useful for exploring a different background, palette, lighting or style.

Choose and upload your sources

Select a model, then check its required image count, maximum image count and file-size limit in the upload area. Use JPG, PNG or WebP files. Models can have different limits; multiple sources are available only where the catalog permits them.

Choose images you have permission to use and make the important subject easy to identify. Uploading several images does not guarantee that every detail will be preserved.

Say what changes and what stays

Preserve the bottle shape, cap and label area. Change the background to dark coastal rocks at sunset. Use warm reflections and subtle water droplets. Keep the product clearly in focus.

For multiple sources, explain the role of each image in its selection order. Keep the instruction focused and inspect the output before making more changes. These prompts are starting drafts, not tested performance claims.

Review model-specific parameters

The creator displays supported resolution, aspect ratio and any output-format, quality or additional runtime controls for the selected model. Legal combinations and Credit costs come from the current catalog. A setting change may adjust other choices; check the notice before submitting.

Preview, download and continue

Open a completed image at full size or download it. Check My Creations for results and status. If the generation did not complete, review its message before starting another request.

Use Text to Image when you want to start from a written idea. To animate a result, download it and add it to Image to Video. All four modes use the same account and Credit balance; current plans appear on Pricing.