Skip to main content
The Edit tool lets you describe global changes to your image (e.g., ‘make the object red’). You can optionally add reference images. There is no brush selection; the AI applies the description to the relevant areas.
Edit tool interface

​
How it works

The AI interprets your text prompt to modify the entire image or specific elements it identifies. Simply describe what you want to change and the AI determines which areas to modify based on your description.

​
Steps

  1. Open your image - Navigate to the image in your Library or board and click to open it in detail view
  2. Access the tool - Click the Edit icon in the vertical toolbar, or access via the Edit section in the left navigation
  3. Describe the change - Enter a clear description of what you want to change (e.g., “make the car blue”, “add clouds to the sky”)
  4. Add reference images (optional) - Upload reference images to guide the style or specific changes
  5. Generate - Click Generate to apply the edit to your image

​
Examples

Popular edits span color adjustments, object modifications, and creative enhancements. Colors - Change clothing, adjust skin tones, modify backgrounds or lighting. Objects - Add accessories, change hairstyles, modify architectural elements. Creative - Apply artistic effects, create seasonal variations, or stylize specific areas.

​
Tips

Tip: Be specific — include lighting and style in your description for accurate results. Tip: Add reference images to guide complex transformations. Tip: Make incremental changes rather than one dramatic edit.

​
Troubleshooting

Unnatural results: Refine your prompt with more specifics, or split complex changes into multiple edits. Lighting inconsistencies: Include lighting context in your prompt (for example, “maintain soft morning light”). Unexpected changes: Specify which elements to modify (for example, “change only the car color”).

​
Edit specific areas with masking (beta)

This is an experimental feature. You might experience a higher number of failures when using this feature.
You can drag and select a rectangular area of the image to apply specific updates to that region only. You can provide reference images along with a prompt before you submit the edit. It is useful for fixing small details that are not reproduced accurately by the AI when generating the original product shot. Find the image that you love and then use this image to make final fixes to get it ready for publishing. Tip: Fix small AI artifacts like strange eyes or blemishes by simply selecting the area and typing a prompt like “Make the eyes look natural” or “Remove the blemish.” Tip: When fixing logos, select the logo area with some padding around it. Add a closeup of the logo as a reference (usually by cropping or taking a screenshot of the logo area on the product). Use a prompt like “Fix the logo on the tshirt.” Tip: If you notice blending issues, try sending a more closely cropped reference with a specific prompt like “Fix the logo and text on the red tag.”
When trying to fix product accuracy issues, make sure to provide a reference area of the product manually cropped. Without it, the AI doesn’t know how to fix the original image.
Do not upload a full image of a product or person as reference.

​
Credit usage

Each edit uses 10 credits per run.

​
See also