Outpainting

A guide to the Outpainting action for expanding images and extending their scenes with AI.

Outpainting is a creative AI technique that extends an image beyond its original borders. It analyzes the existing image and a text prompt to generate new content that seamlessly expands the original scene, allowing you to change an image's aspect ratio, un-crop a photo, or create vast, panoramic landscapes from a small starting picture.

Use this action to:

  • Convert a vertical photo into a horizontal banner for a website.

  • Fix a photo that was framed too tightly by adding more context around the subject.

  • Create artistic, infinitely zoomable images.

How It Works

The Outpainting action takes an input image and effectively places it onto a larger, transparent canvas. You specify how much the canvas should expand in each direction (left, right, top, bottom). The AI then fills in the new, empty areas with generated content that matches the style, lighting, and subject matter of the original image, guided by your text prompt.

Before & After Example: Expanding a Portrait

Original Image
Result (Expanded to Landscape)

Configuration

Input Parameters

Parameter
Type
Description

Image

File

The source image you want to expand.

Prompt

Text

A description of the scene. This guides the AI in generating the new areas. Be descriptive!

Left / Right / Top / Bottom

Number

The number of pixels to expand the canvas in each direction.

Output Parameters

Parameter
Type
Description

Output

File

The final, expanded image.

Example: Creating a Website Banner

In the example above, we started with a portrait-oriented photo and needed a wide banner for a website header.

  1. Input: The original, vertically-oriented photo.

  2. Configure the Outpainting Action:

    • Image: Connect the original photo file.

    • Prompt: A beautiful woman with long hair, standing in a field of wildflowers during a golden sunset, epic landscape, photo-realistic.

    • Left: 300 (pixels)

    • Right: 300 (pixels)

    • Top: 50 (pixels)

    • Bottom: 50 (pixels)

The AI uses the prompt and the existing pixels to dream up what the rest of the scene looks like, generating more wildflowers, sky, and landscape to the left and right of the original subject, effectively changing the image's entire composition and aspect ratio.

Last updated

Was this helpful?