a tightly cropped interior photograph with obvious room beyond the frame; then Extend the photograph laterally with coherent architecture, perspective, and lighting.
After
Before
Generative outpainting

AI Image Extender — Expand Photos Beyond Every Edge

Generated edges can vary; review repeating patterns, faces, text, and architecture

Extend a photo left, right, up, or down without stretching or cropping it. New pixels follow the original scene while the prepared source frame is copied exactly into the larger canvas.

  • Square, portrait, story, and widescreen output
  • Prepared source frame copied exactly after generation
  • Optional direction for scenery, lighting, and composition

Choose the photo you want to expand

JPG, PNG, WebP, HEIC, and other supported formats. Free up to 20 MB; Premium and higher up to 50 MB for this tool.

Once the canvas is wide enough, turn it into a 1280×720 thumbnail with a bold headline; if a crop would do instead, reframe with sliders for position.

Anything odd that appears in the new edges can be erased by describing it.

What you can do with AI Image Extender

Review the main capabilities before you begin.

  • Extend photos beyond their original borders
  • Convert full images to new aspect ratios without cropping
  • Generate matching scenery around the source frame
  • Create portrait, story, square, and widescreen layouts

Settings reference

Available controls for AI Image Extender.

Workflow

Single image or Batch imagesPresets
Process one image, or use Premium Batch images to apply the same automatic recipe to a set and download one ZIP.Options: Single imageBatch images

Layouts

One-tap presetsPresets
Pick a ready-made direction with one tap; generation uses the selected preset automatically.Options: Square expansionWide bannerVertical storyPortrait post
Custom instructions (optional)Text input
Describe the requested change, relevant materials, lighting, perspective, and details that must remain unchanged.

Output

Expanded aspect ratioPresets
Choose the larger output canvas that generated pixels will fill around the complete original image.Options: Square 1:1Portrait 4:5Story 9:16Wide 16:9Portrait 3:4Landscape 4:3
Ask AI to preserve subject identity and product detailsOn / Off
Adds an explicit instruction to retain the main subject, shape, branding, labels, and readable text unless the request changes them.
Free to use · No signup required · Without watermarkUploads are encrypted in transitRemoved under our retention policy

Frequently Asked Questions

Watch each transformation land

Every photo crosses the slash untouched and comes out the other side finished.