Bytedance Seedream V5.5 Layer Decomposition API Documentation

Bytedance Seedream V5.5 Layer Decomposition API Documentation

Playground

Seedream 5.5 Layer Decomposition separates a single image into a base image and transparent layers, enabling flexible compositing, image editing, asset extraction, and layered design workflows. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Seedream 5.5 Layer Decomposition separates a single input image into a base image and transparent image layers for compositing, design, and editing workflows. It breaks a flat image into editable visual components that can be reused, rearranged, or refined in downstream creative tools.

Why Choose This?

• Layered image decomposition
  Separate one image into a base image and multiple transparent layers.

• Transparent PNG layers
  Every layer is returned as a PNG with an alpha channel, ready to drop into a design tool.

• Editing-ready outputs
  Use decomposed layers for compositing, retouching, layout design, and creative editing.

• Prompt-guided layer control
  Optionally describe the desired layers, targets, objects, or grouping behavior.

• Automatic element detection
  When no prompt is provided, the model can identify key visual elements automatically.

• Flexible resolution tiers
  Choose 1k, 1.5k, or 2k depending on your workflow needs.

Parameters

How to Use

1. Upload an image — Provide a single PNG or JPEG image for decomposition.
2. Add a prompt optional — Describe the objects, groups, targets, or layer structure you want.
3. Choose resolution — Select 1k, 1.5k, or 2k.
4. Set max_images optional — A larger value allows a finer split and costs more.
5. Submit — Generate the decomposed image result: the base image first, then the layers.

Pricing

Each request costs max_images times the per-image price of the selected resolution. You are charged for max_images, the most images the request can return, not for the number actually returned: a simple image may come back with fewer layers than max_images and still costs the full amount. Set max_images to the number of layers you need plus one for the base image.

The default max_images is 30, so a default request costs $0.60 at 1k or 1.5k and $1.05 at 2k. Examples: max_images 5 at 1k costs $0.10; max_images 100 at 2k costs $3.50.

Best Use Cases

• Compositing workflows — Separate foreground objects and background elements for layered editing.
• Design production — Convert flat images into editable visual components.
• Marketing asset preparation — Extract reusable objects, people, products, or scene elements from campaign images.
• Creative editing — Rearrange, restyle, replace, or refine isolated image layers.
• Layout workflows — Prepare clean visual parts for presentations, ads, thumbnails, and design systems.

Pro Tips

• Use a clear image with distinct foreground, subject, and background regions.
• Add a prompt when you need specific objects, groupings, or layer behavior, and keep it focused on what should become separate layers.
• Choose 1k or 1.5k for lower-cost iteration, and 2k when higher-resolution layers are needed.
• Avoid cluttered scenes when you need precise layer separation.

Related Models

• Seedream 5.5 — Generate images from text prompts.
• Seedream 5.5 Edit — Edit images with up to 10 references.
• Seedream V5.0 Pro Layer Decomposition — The previous Pro layer decomposition model.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "resolution": "1k",
  "max_images": 30,
  "output_format": "png"
}
JSON
)

# 1. Submit the prediction once. Never retry this request automatically:
# a retry after an unclear failure can create a second, billed task.
CHECK='The task may already exist. Check History or the predictions list before submitting again.'
if ! SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --connect-timeout 10 --max-time 60 \
  -X POST "https://api.wavespeed.ai/api/v3/bytedance/seedream-v5.5/layer-decomposition" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}"); then
  printf 'Submit failed. %s\n' "${CHECK}" >&2
  exit 1
fi

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id. %s\n' "${CHECK}" >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction reaches a final status.
BODY_FILE=$(mktemp)
trap 'rm -f "${BODY_FILE}"' EXIT
while true; do
  # Transient errors (429, 5xx) are retried, honoring Retry-After.
  HTTP_CODE=$(curl --silent --show-error --fail-with-body \
    --connect-timeout 10 --max-time 60 --retry 5 --retry-delay 2 \
    --output "${BODY_FILE}" --write-out '%{http_code}' \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}" || true)
  case "${HTTP_CODE}" in
    2??) ;;
    000|429|5??)  # dropped connection or temporary error: keep polling
      printf 'Task %s: HTTP %s, polling again in 10s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      sleep 10; continue ;;
    *) printf 'Task %s: HTTP %s\n' "${PREDICTION_ID}" "${HTTP_CODE}" >&2
      cat "${BODY_FILE}" >&2; exit 1 ;;
  esac
  RESULT=$(jq 'if type == "object" and has("data") then .data else . end' "${BODY_FILE}")
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout|deleted)
      printf 'Task %s ended with status %s\n' "${PREDICTION_ID}" "${STATUS}" >&2
      printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Every model also accepts the global request options enable_sync_mode and enable_base64_output; see their pages for details.

Response Parameters

Result Request Parameters

Result Response Parameters