AI Image Expander
Extend a photo beyond its edges to make room for a new composition. Upload an image, choose a canvas ratio and position the original inside it. AI fills the added area; review the result before using it.
A product photo expanded for a wider layout
This example adds space on the left and right of a perfume photo. Compare the source with the actual expansion result below. The added area is generated, not recovered from a hidden original.
- Compare the labels and bottle outlines with the source.
- Inspect the joins between the original and generated background.
- Leave copy outside the product and add it in your layout afterward.


Change the composition, not just the pixel count
Image expansion adds new content beyond the frame. It is useful when a crop is too tight for a different layout. Increasing resolution with an upscaler keeps the same composition instead.
Product banners
Extend the surroundings of a product photo for a horizontal layout. Keep the product inside the original area and use the added space for copy in your design.
Portrait and social layouts
Place an existing photo inside a taller canvas, such as 9:16. Review new scenery and any generated parts of a person before publishing.
Reusable campaign images
Adapt one source to a square, portrait or landscape composition. Save the original separately and compare each expanded version before using it.
How to expand a photo in Kivue
Upload the source
Choose a PNG, JPG or WebP image up to 5 MB. Each side must be at least 64 pixels and no larger than 4,090 pixels.
Set the canvas
Choose a ratio and position or scale the photo in the editor. The canvas must extend beyond the original; selecting a ratio alone does not add space if it already fits.
Generate and inspect
Review the displayed credit cost, start the task and compare the completed image with the source. Download the result or continue editing from its controls.
Image limits and canvas ratios
Larger rotated or expanded canvases can exceed the processing limit even when the source is accepted. Reduce the expansion range if the editor reports a capacity limit. AI-generated areas can contain invented objects, incorrect anatomy or inconsistent text.
Available ratio presets
1:1 · 3:2 · 4:3 · 5:4 · 16:9 · 21:9 · 2:3 · 3:4 · 4:5 · 9:16
- Input formats
- PNG, JPG and WebP
- Upload limit
- 5 MB per image
- Input dimensions
- 64–4,090 px on each side
- Processing cost
- 10 credits per image
Expand, upscale or crop an image?
Each operation changes a different part of the image. Choose by the result you need.
| Operation | What it changes | Use it for |
|---|---|---|
| Image Expander | Add generated content beyond the edges. | More space around the subject |
| Image Upscaler | Increase resolution within the existing frame. | A larger version of the composition |
| Cropping | Remove content from the edges in an editor. | A tighter composition; no new scenery |
Frequently asked questions
How do I expand a product photo without cropping the product?
Upload the photo, choose a larger canvas and keep the product inside the original image area. Adjust its position in the editor, then generate the added surroundings. Compare product labels and edges with the source afterward.
Can I turn a landscape photo into a vertical 9:16 image?
Yes. Choose 9:16 and position the original in the taller canvas. The tool generates the added area above or below it. New scenery and body details are predictions, so inspect them before publishing.
Is AI image expansion the same as image upscaling?
No. Expansion generates content outside the original frame and can change its aspect ratio. Upscaling increases resolution within the existing composition. Use the Image Upscaler when you need more pixels without additional scenery.
Which file sizes and dimensions does the image expander accept?
Use PNG, JPG or WebP files up to 5 MB. Each side must be at least 64 pixels and no larger than 4,090 pixels. WebP is converted before processing. Extra-large expanded or rotated canvases may need a smaller expansion range.
Will AI recover the exact scene outside the original photo?
No. The model invents surroundings using the visible image as context. It cannot recover unseen details. Treat expanded areas as generated content and check product identity, patterns, text and perspective against the source.
How many credits does it cost to expand an image?
The tool displays the current credit cost before submission. Check that amount and the selected canvas before you start. This guide uses the live tool configuration for its displayed credit amount rather than a separate price list.
Give your photo room for a new layout
Choose a canvas, position the source and review the expanded edges.