Upload a reference image, describe the change you want, and create edited still assets for video references, thumbnails, products, and campaign visuals.
Start from an uploaded image and describe the edit or refinement you need.
Choose 1K, 2K, or 4K output in the image generator before running.
Use source images when the subject, product, pose, or composition matters.
GPT Image 2 Edit is the image editing path available in the seedance 2.1 image workspace.
Use GPT Image 2 Edit when a task starts from an uploaded reference image instead of only a text prompt.
The image generator keeps reference upload, prompt, resolution, output format, and credit preview visible before you create.
Creative engine
Upload a source image, describe the edit, choose output settings, then preview credits before generating.
Use these directions to make the source image, requested change, and final use clear before spending credits.

Describe what should stay recognizable and what background, lighting, or mood should change.
Image editing is useful when you already have a reference and need a controlled still asset.
Start from an uploaded image when subject or product continuity matters.
Choose 1K, 2K, or 4K depending on the asset you need.
The generator previews credits based on quality, resolution, and image count.
Prepare cleaner reference frames before writing an image-to-video prompt.
Practical controls for editing reference images in the seedance 2.1 workspace.
Upload an image and describe the edit you want.
Use image mode when the task depends on a source image.
Choose PNG or JPEG output format in the image generator.
GPT Image 2 Edit is available from image-to-image mode with an uploaded reference image.
A quick summary of the options available in the image generator.
Use GPT Image 2 Edit when you want to refine an uploaded image.
Start from image mode with GPT Image 2 Edit selected.
Add the source image that should guide the edit.
Write what should change, what should stay, and where the final image will be used.
Preview credits, create the image, then use it as a still asset or video reference.
Use GPT Image 2 Edit for still-image tasks that need a source image.
Refine a still frame before using it in an image-to-video prompt.
Adapt product visuals, backgrounds, and presentation styles from a source image.
Adjust a generated or uploaded still into a stronger thumbnail candidate.
Use an existing image as the anchor while changing mood, lighting, or art direction.
Questions about GPT Image 2 Edit in the seedance 2.1 workspace.
Open image mode with GPT Image 2 Edit selected and preview credits before creating.
Cinematic Seedance video path with text and image prompts, prompt ideas, and credit preview.
Standard Seedance video mode for 480p, 720p, and 1080p generation.
A faster, lower-cost Seedance video option for 480p and 720p drafts.
Create reference images, style frames, thumbnails, and product visuals for image or video projects.