Vidu Start End To Video 2.0 API Documentation

Vidu Start End To Video 2.0 API Documentation

Playground

Vidu Start-End to Video 2.0 generates smooth transition videos interpolating between given start and end images for natural morphing effects. Ready-to-use REST API, best performance, no coldstarts, affordable pricing.

Features

Vidu Start-End-to-Video 2.0 generates a short video that transitions from a start image to an end image, guided by a text prompt. You provide the first frame (image) and the final target frame (last_image), then describe the transformation story in the prompt. The model fills in the in-between motion and visual continuity, making it ideal for morphs, product reveals, character transformations, and “before → after” storytelling. You can control how dramatic the motion is with movement_amplitude.

Key capabilities

• Start-to-end guided video generation (image → last_image)
• Prompt-driven transformation with controllable motion intensity
• Designed for smooth visual transitions and narrative continuity
• Movement amplitude control: auto / small / medium / large
• Seed control for reproducible variations

Use cases

• Character transformation and morph sequences (human → creature, outfit swaps, style shifts)
• Product reveals (closed box → unboxed product, sketch → final render)
• Before/after transitions (room makeover, color grading shift, weather change)
• Logo or brand animations (static mark → dynamic finish frame)
• Short cinematic transitions for ads and social clips

Pricing

Inputs

• prompt (required): describe the transition and what happens in-between
• image (required): start image (first frame)
• last_image (required): end image (final frame)

Parameters

• movement_amplitude: motion intensity (auto, small, medium, large)
• seed: random seed (-1 for random; fixed for reproducible results)

Prompting guide (start → end)

Write prompts that explain how the start becomes the end:

• Describe the transformation mechanism (morphing, unfolding, dissolving, assembling)
• Call out key beats (sparks, parts shifting, panels opening, smoke/energy)
• Mention camera intent (fixed shot, slow push-in) if you need it
• Keep continuity constraints clear (same framing, stable background, consistent lighting)

Template: Fixed camera. Start as [start]. The subject transforms by [mechanism] into [end]. Smooth transition, coherent motion, consistent lighting, no flicker.

Example prompts

• A metal suit transforms into a sleek sports car through mechanical folding panels, glowing arc energy, parts assembling smoothly, fixed camera, cinematic lighting, seamless morph from start to end.
• A sketch drawing gradually fills with color and detail, turning into a polished illustration, gentle dust particles, slow push-in, smooth transition.
• A winter landscape melts into spring: snow dissolves, grass emerges, flowers bloom, warm sunlight grows stronger, smooth continuous motion.

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'
{
  "prompt": "A cinematic ocean wave at sunrise, highly detailed",
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "last_image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "movement_amplitude": "auto",
  "seed": -1
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/vidu/start-end-to-video-2.0" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

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 "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    created|processing) sleep 2 ;;
    *) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters