Bytedance Seedance 2.5 Image To Video API Documentation

Bytedance Seedance 2.5 Image To Video API Documentation

Playground

Seedance 2.5 (Image-to-Video) generates Hollywood-grade cinematic videos from reference images and text prompts with native audio-visual synchronization, director-level camera and lighting control, and exceptional motion stability. Built on Seed’s unified multimodal architecture, it preserves the input image’s subject and composition while adding expressive, physically accurate motion.

Features

Seedance 2.5 Image-to-Video generates cinematic videos from a reference image and text prompt. It preserves the input image’s subject, composition, and style while adding expressive motion, camera movement, scene progression, and synchronized audio.

Why Choose This?

• Image-to-video generation
  Animate a reference image into a cinematic video with prompt-guided motion.

• Image-faithful results
  Preserve the input image’s subject identity, composition, lighting, and style while generating motion.

• Optional last-frame guidance
  Use last_image to guide the final frame or ending direction of the generated video.

• Native audio generation
  Generate synchronized audio together with the video output.

• Director-style prompt control
  Control camera movement, lighting, mood, subject action, and scene progression through natural-language prompts.

• High-resolution output
  Generate videos in 480p, 720p, 1080p, or 4k.

Parameters

The output aspect ratio follows the input image automatically.

How to Use

1. Upload a start image — Provide the image that should guide the first frame, subject, composition, and style.
2. Write your prompt — Describe the motion, action, camera movement, lighting, mood, and scene progression.
3. Add last image optional — Use last_image when you want to guide the final frame or ending direction.
4. Set duration — Choose a video length from 4 to 30 seconds.
5. Choose resolution — Select 480p, 720p, 1080p, or 4k.
6. Configure audio optional — Keep generate_audio enabled when synchronized audio is needed.
7. Submit — Generate the final image-to-video output.

Writing Effective Prompts

The input image sets the first frame — your prompt directs what happens from there. The clearer the motion and camera direction, the better the result.

1. Start from the frame

Describe the action that begins from your image: what the subject does, where it goes, how the camera moves. Don’t re-describe what’s already visible in the image — build on it.

The woman in the photo turns from the window and walks toward the camera; slow push-in, warm interior light.

2. Shot-by-shot timeline

For longer clips, break the motion into timed segments (1-second granularity), continuous, no gaps.

> 0-2s: She turns from the window, a faint smile forming. Camera holds. > 2-5s: She walks toward the lens and stops mid-frame; slow push-in, shallow focus. > 5-8s: She glances off-screen as light shifts from warm to cool. Handheld drift.

3. Consistency notes

Lock the look for the whole clip: camera style, lighting, palette, atmosphere, audio.

Camera language

Write these directly; the model understands them:

• Shot size — extreme wide, wide, medium, medium close-up, close-up
• Movement — push in, pull out, pan, track, follow, orbit, tilt up, handheld shake
• Angle — low angle, overhead, eye-level, first-person
• Techniques — one-shot / long take, dolly zoom, bullet time, speed ramp

For a niche term, add a plain-language gloss: “rack focus: the sharp foreground softens as the background comes into focus.”

Referencing uploaded assets

When you attach images, videos, or audio, bind each one explicitly in the prompt by its upload order — @image1, @video1, @audio1 — and say what it’s for. Don’t rely on labels drawn inside the image itself.

• “The knight in @image1 walks through the castle in @image2.”
• “Refer to @video1 for the camera movement only; keep its shot order.”
• “@image1 and @image2 are Character 1, voiced by @audio1.”

When a reference is already accurate, just point to it — no need to re-describe it in detail.

Negative control

Positive descriptions work best, but you can suppress subtitles and audio:

• “No subtitles.”
• “No BGM — ambient and action sounds only.”
• “No audio.”

Weak vs. strong

Pricing

Pricing is based on selected resolution and duration.

Example Costs

Best Use Cases

• Product demos — Animate product shots into cinematic showcase videos.
• Ad creatives — Turn storyboard frames or campaign stills into polished commercial footage.
• Character animation — Bring character art, portraits, or illustrated subjects to life with natural motion.
• Scene development — Transform a single keyframe into a full cinematic sequence.
• Start-and-end-frame generation — Use image and last_image to guide both the beginning and ending of the video.
• Social media content — Generate short-form videos from still images for vertical, square, or landscape-style layouts.

Pro Tips

• Upload high-quality reference images for stronger subject preservation.
• Write prompts like a film director, including action, camera movement, lighting, mood, and atmosphere.
• Use last_image when the ending frame or final visual direction matters.
• Start with a short duration for quick iteration, then increase duration once the motion direction is confirmed.
• Use 480p or 720p for testing and 1080p or 4k for higher-resolution output.
• Describe character expressions and actions for more engaging motion.

Related Models

• Seedance 2.5 Text-to-Video — Generate video from text prompts alone.
• Seedance 2.5 Video Edit — Edit an existing video with a prompt.
• Seedance 2.5 Video Extend — Extend an existing video with a new segment.
• Seedance 2.0 Image-to-Video — Previous-generation Seedance image-to-video 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'
{
  "prompt": "A cinematic ocean wave at sunrise, highly detailed",
  "image": "https://interactive-examples.mdn.mozilla.net/media/cc0-images/painted-hand-298-332.jpg",
  "resolution": "720p",
  "duration": 5,
  "generate_audio": true
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/bytedance/seedance-2.5/image-to-video" \
  -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 "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 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|deleted) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters