Seedance 2.0 API

中文文档

Generate AI videos from text, images, audio, and video references — powered by ByteDance Seedance 2.0.

Features

• Text to Video — describe a scene, get a video
• Image to Video — animate any still image
• Video Reference — transfer motion/choreography from a reference video
• Audio Sync — synchronize video to music or voiceover
• Image + Audio — lip-sync a portrait with speech audio
• 4–15 seconds output duration, up to 1080p
• Two model tiers: seedance_2.0_fast (budget) and seedance_2.0 (premium)

Quick Start

1. Get Your API Key

Sign up at seed2.io and copy your API key from the Account page. New users get 10,000 free credits (≈ $1).

2. Generate a Video

cURL:

curl -X POST https://api.seed2.io/v1/video/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A cat playing piano in a jazz bar, cinematic lighting",
    "model": "seedance_2.0_fast",
    "duration": 5,
    "ratio": "16:9"
  }'

Python (with SDK):

from seedance import Seedance

client = Seedance(api_key="sk-s2-xxxx")
result = client.generate(
    prompt="A cat playing piano in a jazz bar, cinematic lighting",
    model="seedance_2.0_fast",
    duration=5,
)
print(result["video_url"])

Node.js:

const res = await fetch("https://api.seed2.io/v1/video/generate", {
  method: "POST",
  headers: {
    Authorization: "Bearer YOUR_API_KEY",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    prompt: "A cat playing piano in a jazz bar, cinematic lighting",
    model: "seedance_2.0_fast",
    duration: 5,
    ratio: "16:9",
  }),
});
const { data } = await res.json();
console.log("Task ID:", data.task_id);

3. Poll for Result

curl https://api.seed2.io/v1/tasks/TASK_ID \
  -H "Authorization: Bearer YOUR_API_KEY"

Response when completed:

{
  "code": 0,
  "message": "ok",
  "data": {
    "id": 123,
    "status": "completed",
    "result": {
      "video_url": "https://cdn.seed2.io/videos/xxx.mp4",
      "duration": 5
    },
    "points_cost": 2900
  }
}

API Reference

Authentication

All requests require an API key in the Authorization header:

Authorization: Bearer sk-s2-xxxxxxxxxxxx

Create Video Task

POST /v1/video/generate

Parameter	Type	Required	Description
prompt	string	Yes	Text description. Use @image_file_1 to reference uploaded images.
model	string	No	seedance_2.0_fast (default) or seedance_2.0
duration	int	No	4–15 seconds. Default: 5
ratio	string	No	16:9, 9:16, 1:1, 4:3, 3:4, 21:9. Default: 16:9
image_files	string[]	No	Image URLs (max 9). Reference in prompt as @image_file_1, @image_file_2, etc.
video_files	string[]	No	Reference video URLs (max 3, total ≤ 15s)
audio_files	string[]	No	Audio URLs (max 3)

Response:

{
  "code": 0,
  "message": "ok",
  "data": {
    "task_id": 456,
    "status": "pending",
    "points_cost": 2900,
    "balance": 7100
  }
}

Query Task

GET /v1/tasks/:task_id

Status values: pending → processing → completed / failed

If a task fails, credits are automatically refunded.

Usage Modes

Text to Video

The simplest mode — just provide a prompt.

result = client.generate(prompt="Sunrise over Mount Fuji, timelapse, 4K")

Image to Video

Animate a still image. Reference images in the prompt with @image_file_N.

result = client.generate(
    prompt="The woman in @image_file_1 walks through a garden, petals falling",
    image_files=["https://example.com/woman.jpg"],
)

Video Reference

Use a reference video to transfer motion, style, or choreography.

result = client.generate(
    prompt="A cartoon character performing the same dance",
    model="seedance_2.0",
    duration=10,
    video_files=["https://example.com/dance.mp4"],
)

Audio Sync

Synchronize the generated video with background music or sound effects.

result = client.generate(
    prompt="A DJ mixing at a nightclub, energetic crowd",
    duration=10,
    audio_files=["https://example.com/edm-track.mp3"],
)

Image + Audio (Lip-sync)

Combine a portrait with speech audio for realistic lip-sync.

result = client.generate(
    prompt="The person in @image_file_1 is giving a TED talk with confident gestures",
    model="seedance_2.0",
    duration=8,
    ratio="9:16",
    image_files=["https://example.com/speaker.jpg"],
    audio_files=["https://example.com/speech.mp3"],
)

Pricing

Model	Without Video/Audio	With Video/Audio
seedance_2.0_fast	580 credits/sec	1,160 credits/sec
seedance_2.0	1,160 credits/sec	2,320 credits/sec

• 1 USD = 10,000 credits
• New users: 10,000 free credits
• Recharge at seed2.io/recharge

Example cost: A 5-second text-to-video with seedance_2.0_fast costs 2,900 credits ($0.29).

Examples

Check the examples/ directory for ready-to-run scripts:

File	Language	Description
examples/python/text_to_video.py	Python	Text to Video
examples/python/image_to_video.py	Python	Image to Video
examples/python/video_reference.py	Python	Video Reference
examples/python/image_audio_lipsync.py	Python	Image + Audio Lip-sync
examples/python/batch_generate.py	Python	Batch Generation
examples/javascript/text_to_video.mjs	Node.js	Text to Video
examples/javascript/image_to_video.mjs	Node.js	Image to Video
examples/curl/examples.sh	cURL	All modes

Run any Python example:

export SEEDANCE_API_KEY="sk-s2-xxxx"
python examples/python/text_to_video.py

Python SDK

This repo includes a lightweight Python client in seedance/:

from seedance import Seedance

client = Seedance(api_key="sk-s2-xxxx")

# One-liner: create → poll → return result
result = client.generate(prompt="A cat playing piano")
print(result["video_url"])

# Or manage tasks manually
data = client.create_task(prompt="A cat playing piano")
task = client.wait_for_task(data["task_id"])
print(task["result"]["video_url"])

Requirements: Python 3.7+ with requests installed (pip install requests).

Links

• Web App: seed2.io
• API Docs: seed2.io/api-docs
• Video Gallery: seed2.io/plaza

License

MIT