AI Image Editor for Agents and People

Image Editor

Retouch

Place Object

Resize for Platforms

Crop

Adjust

Paint over the area to change. Use [ and ] to resize the brush; press Escape to clear.

Resize for where you sell

Pick a platform and use. We handle the exact publishing size.

Destination and size

Platform

Select a platform and use to see the exact size.

Canvas background (for margins)

Used when the product ratio differs from the target size. Amazon recommends pure white.

3 How should the image fit? AI smart canvas Up to 2 Credits Best finish when the target shape changes. Locks your product, text and logo; AI fills only new margins. Keep everything (Fit) Free · Recommended Preserves the full product without cropping; adds margins in your chosen background color. Fill canvas (Cover) Free Fills the canvas without margins; outer edges may be slightly cropped to fit.

Keep everything is the safest choice for product listings.

Choose a size to check availability and exact credits.

Want to preview all 3 variations at once?

Upload Image

Drag drop or click to upload

JPG, PNG, WEBP up to 10MB

Agent-first and human-friendly

A composable AI image editor for agents and people

Use visual browser controls for hands-on editing or let an agent call structured inpaint, adjust, filter, place, and crop operations against owned SceneOf assets.

Inpaint, adjust, filter, place, and crop
Masks and owned asset inputs
Credit-free local crop; quoted AI edits

Connect an AI agent

MCP tool

edit_image

Copy this prompt to your agent

Edit an image using SceneOf MCP.

Tool: edit_image
Capabilities: inpaint, adjust, filter, place, crop

Setup (if not already connected):
  Connection guide: https://sceneof.com/mcp/docs/connect.md

Instructions:
1. Connect to SceneOf MCP if not already connected.
2. Choose an owned image asset or provide inline image data.
3. Describe the edit — add a mask or object image when required.
4. Call edit_image and verify the returned asset before reuse.
5. Report the final credit charge and saved asset URL.

How the agent workflow works

1. Choose an owned image or provide inline image data.
2. Describe the edit and add a mask/object image when required.
3. Run edit_image and verify the returned asset before reuse.

Frequently asked questions

Can an agent edit an existing SceneOf image?

Yes. Pass the image asset ID to avoid uploading it again.

Which edits need a mask?

Inpainting needs a transparent PNG mask matching the source dimensions.

Are edits reversible?

The source asset remains available and the edited result is stored as a new image.