Home/Explore/Image Editing/bytedance/seededit-v3
image-to-image

image-to-image

Bytedance SeedEdit V3

bytedance/seededit-v3

SeedEdit 3.0 by Bytedance enables prompt-guided image editing for precise object, scene, and layout changes without disrupting composition. Ready-to-use REST API, high performance, no coldstarts, affordable pricing.

Hint: You can drag and drop a file or click to upload

preview
If enabled, the output will be encoded into a BASE64 string instead of a URL. This property is only available through the API.

Idle

Rainbow colored short curly hair

Your request will cost $0.027 per run.

For $1 you can run this model approximately 37 times.

One more thing::

ExamplesView all

Rainbow colored short curly hair
"A bright yellow cartoon school bus arriving at a jungle bus stop. The bear and the monkey in matching school uniforms (navy blazer, white shirt, gray shorts, caps with 'B' and 'M') prepare to board. The monkey looks confident, the bear looks clumsy. Background with jungle trees and vines."
Become an Anime Dog.
Become cyberpunk style.
Change "love" to "seed"
Turn the scene into daytime
The cat is in its owner's arms.
Turn the bubble into a heart shape.
Remove all pedestrians except the one in the middle.
Make the girl look realistic.

README

SeedEdit v3 — bytedance/seededit-v3

SeedEdit v3 is a prompt-based image editing model for fast, high-quality transformations. Provide a source image and a short instruction, and the model performs targeted edits—such as changing hair, clothing, colors, backgrounds, or adding/removing elements—while keeping the overall composition and identity stable.

Key capabilities

  • Prompt-driven image editing from a single input image
  • Strong preservation of subject identity and scene structure for everyday edits
  • Works well for localized edits (hair, face details, clothing) and global adjustments (color/lighting)
  • Efficient for rapid iteration with controllable randomness via seed

Use cases

  • Portrait edits: hair color/style changes, light makeup tweaks, accessory edits
  • E-commerce variants: recolor outfits/products, background cleanup, consistent product shots
  • Marketing creatives: quick visual iterations (color mood, props, text-free adjustments)
  • Photo cleanup: remove small distractions, fix minor inconsistencies, improve overall clarity
  • Content localization: adapt visual details (e.g., wardrobe color/style) while keeping the same subject

Pricing

OutputPrice
Per image$0.027

Inputs

  • image (required): the source image to edit
  • prompt (required): the edit instruction

Parameters

  • prompt: what to change (and optionally what to keep)
  • image: input image (upload or URL)
  • guidance_scale: how strongly the edit follows the prompt (higher = stronger edit)
  • seed: random seed (-1 for random; fixed value for reproducible results)

Prompting guide

For clean edits, be explicit about what must stay the same:

Template: Keep [identity/pose/background]. Change [target]. Maintain [lighting/style realism].

Example prompts

  • Keep the face and pose unchanged. Change the hair to short, curly, rainbow-colored hair. Keep natural lighting and realistic texture.
  • Keep the subject identity and background unchanged. Change the jacket color to black and add subtle fabric texture.
  • Remove the object in the background and keep the subject sharp and unchanged.