Prepare a Stable Diffusion image output

A finished file. A reviewed copy.

Prepare the exact asset you intend to use.

1. Choose the Stable Diffusion file

   Keep your original workflow and output file if you need reproducible settings. Use the PNG AI cleaner for supported generation records.

2. Review what changes

   Read the supported findings before cleaning. A full clean removes supported camera, location, editing and provenance records. Use GPS-only cleaning when you need to retain other camera settings.

3. Inspect the downloaded copy

   A broad photo clean removes more than generation fields; compare the policy with the records you want to keep.

Know the output before you start.

This workspace accepts JPG, PNG, and WebP still images. Up to 30 files, 15 MB each, and no more than 50 megapixels. Animated PNG and WebP are rejected.

JPG, PNG, and WebP stay in the same format. The copy is re-encoded, so compression and color can change. A HEIC still uses the separate HEIC cleaner. A video uses the video cleaner. This workspace does not extract frames from a video.

JPG, PNG, WebP

New copy in the same image format

HEIC / HEIF

Use the HEIC cleaner

MP4 / MOV

Use the separate video workspace

An illustrative workflow

stable-diffusion-scene.png

Keep your original workflow and output file if you need reproducible settings. Use the PNG AI cleaner for supported generation records.

Example only. No file has been scanned until you choose one.

Source

Your finished Stable Diffusion still image

Action

Inspect, then create a full metadata-cleaned copy

Review

A broad photo clean removes more than generation fields; compare the policy with the records you want to keep.

Keep

The original and any project or attribution record you need

Before your next export or upload.

Answers for this export workflow.