개발자 API

SeeDanceE2.ai 기반으로 구축하세요.

기반이 되는 Seedance 2.0 모델을 바탕으로 상위 제공업체의 기능을 통합해 캡슐화하고, 표준화된 공개 API로 제공합니다.

model: promodel: fastmodel: mini

워크스페이스 열기 엔드포인트 레퍼런스 읽기

안정적인 API 계약

`pro`, `fast`, `mini`를 한 번만 통합하면 상위 제공업체의 변경 사항은 동일한 공개 계약 뒤에서 흡수됩니다.

텍스트 및 이미지 워크플로

프롬프트 우선 및 이미지 우선 비디오 생성을 하나의 API 표면에서 실행하고, 상위 제공업체별 차이를 고객에게 노출하지 않습니다.

비동기 작업, 폴링, 웹훅

생성 요청을 제출하고 작업 상태를 폴링하거나, 비디오가 준비되면 깔끔한 SeeDanceE2 응답 형태로 웹훅을 받을 수 있습니다.

Public contract

One API surface for video generation

The public API keeps upstream providers hidden. Developers integrate with three public model names, one auth pattern, and one async job lifecycle.

Authentication

Bearer API keys

Send `Authorization: Bearer sk_seedancee2_live_xxxxx` with every request.

Task lifecycle

Submit, poll, webhook

Create an async generation task, poll by ID, or receive completion callbacks.

Usage and credits

Reserved and charged

Responses expose reserved, charged, and refunded credits so billing stays auditable.

Public model names

Use only `pro`, `fast`, and `mini`

pro

Production quality

Use for customer-facing clips and standard production work.

480p, 720p, 1080p

fast

Faster iteration

Use for preview workflows and repeated prompt experiments.

480p, 720p

mini

Lower-cost tests

Use for lightweight drafts and budget-sensitive generation.

480p, 720p

API questions

Frequently asked questions

How do I authenticate requests to the video generation API?

Create an API key in your workspace and send it as a bearer token: Authorization: Bearer sk_seedancee2_live_xxxxx on every request.

Is the API synchronous or asynchronous?

Generation is asynchronous. You submit a task, receive a task ID, then poll the status endpoint or register a webhook to be notified when the video is ready.

Which model names can I request?

Only the public tiers pro, fast and mini. The platform routes each request to the best available underlying video model, so your integration never breaks when providers change.

How does API billing work?

API usage draws from the same credit balance as the web app. Responses expose credits_reserved, credits_charged and credits_refunded so every call stays auditable.

Can the API do image-to-video as well as text-to-video?

Yes. Submit a text prompt for text-to-video, or attach a reference image URL for image-to-video; the image composition is preserved and only the described motion is generated.

SeeDanceE2 Public Video API

SeeDanceE2 exposes a stable public API for video generation. Integrations should use the public model names pro, fast, and mini; upstream provider details stay behind the SeeDanceE2 contract.

Base URL

https://api.SeeDanceE2.ai

Quickstart

Get an API key from the SeeDanceE2 team.
Create a video generation task.
Poll the task or receive a webhook when it completes.

Minimal curl example:

curl -X POST "https://api.SeeDanceE2.ai/v1/videos/generations" \
  -H "Authorization: Bearer sk_seedancee2_live_xxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mini",
    "input": {
      "prompt": "A cinematic drone shot of waves breaking against black volcanic rocks",
      "duration": 5,
      "resolution": "720p",
      "aspect_ratio": "16:9"
    },
    "metadata": {
      "customer_job_id": "job_123"
    }
  }'

Minimal JavaScript example:

const response = await fetch("https://api.SeeDanceE2.ai/v1/videos/generations", {
  method: "POST",
  headers: {
    Authorization: "Bearer sk_seedancee2_live_xxxxx",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "mini",
    input: {
      prompt: "A cinematic drone shot of waves breaking against black volcanic rocks",
      duration: 5,
      resolution: "720p",
      aspect_ratio: "16:9",
    },
    metadata: {
      customer_job_id: "job_123",
    },
  }),
});

const result = await response.json();
console.log(result);

Minimal Python example:

import requests

response = requests.post(
    "https://api.SeeDanceE2.ai/v1/videos/generations",
    headers={
        "Authorization": "Bearer sk_seedancee2_live_xxxxx",
        "Content-Type": "application/json",
    },
    json={
        "model": "mini",
        "input": {
            "prompt": "A cinematic drone shot of waves breaking against black volcanic rocks",
            "duration": 5,
            "resolution": "720p",
            "aspect_ratio": "16:9",
        },
        "metadata": {
            "customer_job_id": "job_123",
        },
    },
    timeout=60,
)

print(response.json())

Authentication

All requests use bearer token authentication.

Authorization: Bearer sk_seedancee2_live_xxxxx
Content-Type: application/json

You can also send the key in x-api-key, but Authorization: Bearer ... is recommended.

API keys and model access

Your SeeDanceE2 API key is issued by the SeeDanceE2 team or from your profile.
The same SeeDanceE2 API key works across public models: pro, fast, and mini.
You do not need separate customer-facing API keys for Mini, Fast, or Pro.
Upstream provider credentials are managed by SeeDanceE2 and are never required in customer requests.

If you already have a working SeeDanceE2 API key, switching models only requires changing the request model field. Example:

{
  "model": "mini"
}

to:

{
  "model": "fast"
}

or:

{
  "model": "pro"
}

Public Models

These are the public model names your integration should use.

Model	Positioning	Duration	Resolution
`pro`	Standard quality production generation	4-15s	480p, 720p, 1080p
`fast`	Faster iteration and preview workflows	4-15s	480p, 720p
`mini`	Lower-cost generation	4-15s	480p, 720p

New integrations should use pro, fast, and mini.

Supported aspect ratios:

16:9, 4:3, 1:1, 3:4, 9:16, 21:9, adaptive

Credits

Public API credits follow the same deduction rules visible in the SeeDanceE2 product.

Formula:

credits = duration_seconds * per_second_rate * number_of_outputs

Reference images, videos, and audio currently do not add extra customer-facing surcharges. The reserved and charged credit values in the API response are the source of truth for each task.

Per-second rates currently aligned with the live SeeDanceE2 product billing rules:

Model	480p	720p	1080p
`pro`	10 credits/s	20 credits/s	50 credits/s
`fast`	8 credits/s	16 credits/s	Not supported
`mini`	5 credits/s	10 credits/s	Not supported

Examples:

Scenario	Credits
`mini`, 5s, 720p	50
`pro`, 4s, 480p	40
`fast`, 10s, 720p	160
`pro`, 5s, 720p, 2 reference images	100

If a task fails and the job is eligible for refund, the returned status payload will show the refunded credits.

Task Lifecycle

Task statuses returned by the public API:

Status	Meaning
`pending`	Accepted and waiting to start
`processing`	Running
`completed`	Finished successfully
`failed`	Finished without usable output

Create Video Task

POST /v1/videos/generations

Top-level request fields:

Field	Type	Required	Description
`model`	string	Yes	Public model name: `pro`, `fast`, or `mini`
`input`	object	Yes	Generation parameters
`callback_url`	string	No	Public webhook URL for completion or failure callbacks
`metadata`	object	No	Customer-defined metadata echoed back in polling and webhook responses

Input fields:

Field	Type	Required	Description
`prompt`	string	Yes	Text prompt
`generation_type`	string	No	`omni_reference` or `first_and_last_frames`; default `omni_reference`
`image_urls`	string[]	No	Reference image URLs
`video_url`	string	No	Single reference video URL
`video_urls`	string[]	No	Reference video URLs. Do not send together with `video_url`
`audio_urls`	string[]	No	Reference audio URLs
`start_image_url`	string	No	First frame image for `first_and_last_frames`
`end_image_url`	string	No	Last frame image for `first_and_last_frames`
`duration`	integer	No	Output duration in seconds, 4-15; default 5
`resolution`	string	No	`480p`, `720p`, or `1080p`; available resolutions depend on the selected model
`aspect_ratio`	string	No	`16:9`, `4:3`, `1:1`, `3:4`, `9:16`, `21:9`, or `adaptive`
`generate_audio`	boolean	No	Generate synchronized audio when supported
`watermark`	boolean	No	Request watermarked output
`seed`	integer	No	Deterministic seed
`return_last_frame`	boolean	No	Include the last frame image in the output payload

Full example:

curl -X POST "https://api.SeeDanceE2.ai/v1/videos/generations" \
  -H "Authorization: Bearer sk_seedancee2_live_xxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "pro",
    "callback_url": "https://example.com/webhooks/seedancee2",
    "input": {
      "prompt": "A cinematic tracking shot through a rain-soaked neon market at night",
      "image_urls": [
        "https://example-cdn.com/reference-1.png",
        "https://example-cdn.com/reference-2.png"
      ],
      "duration": 5,
      "resolution": "720p",
      "aspect_ratio": "16:9",
      "generate_audio": true,
      "return_last_frame": true
    },
    "metadata": {
      "customer_job_id": "job_123",
      "campaign": "launch_week"
    }
  }'

Success response:

{
  "data": {
    "id": "vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
    "status": "pending",
    "model": "pro",
    "credits_reserved": 100,
    "created_at": "2026-06-29T10:00:00.000Z"
  }
}

Get Video Task

GET /v1/videos/generations/{id}

Polling example:

const response = await fetch(
  "https://api.SeeDanceE2.ai/v1/videos/generations/vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
  {
    headers: {
      Authorization: "Bearer sk_seedancee2_live_xxxxx",
    },
  }
);

const result = await response.json();
console.log(result);

Processing response:

{
  "data": {
    "id": "vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
    "status": "processing",
    "model": "pro",
    "created_at": "2026-06-29T10:00:00.000Z",
    "metadata": {
      "customer_job_id": "job_123",
      "campaign": "launch_week"
    }
  }
}

Completed response:

{
  "data": {
    "id": "vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
    "status": "completed",
    "model": "pro",
    "created_at": "2026-06-29T10:00:00.000Z",
    "completed_at": "2026-06-29T10:06:20.000Z",
    "credits_charged": 100,
    "output": {
      "video_url": "https://media.SeeDanceE2.ai/videos/2026/06/29/vid_01jz8b7q4x8k2p9w3m5c6n7r8s.mp4",
      "thumbnail_url": "https://media.SeeDanceE2.ai/video-posters/2026/06/29/vid_01jz8b7q4x8k2p9w3m5c6n7r8s.jpg",
      "last_frame_url": "https://media.SeeDanceE2.ai/video-posters/2026/06/29/vid_01jz8b7q4x8k2p9w3m5c6n7r8s.jpg"
    },
    "metadata": {
      "customer_job_id": "job_123",
      "campaign": "launch_week"
    }
  }
}

Failed response:

{
  "data": {
    "id": "vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
    "status": "failed",
    "model": "pro",
    "created_at": "2026-06-29T10:00:00.000Z",
    "credits_charged": 0,
    "credits_refunded": 100,
    "error": {
      "code": "generation_failed",
      "message": "The video could not be generated. Please try again with a different prompt.",
      "type": "generation_error"
    },
    "metadata": {
      "customer_job_id": "job_123",
      "campaign": "launch_week"
    }
  }
}

GET /v1/videos/generations

Supported query parameters:

Field	Type	Description
`status`	string	Filter by `pending`, `processing`, `completed`, or `failed`
`limit`	integer	Max results. Default 20, max 100

Example:

curl "https://api.SeeDanceE2.ai/v1/videos/generations?status=completed&limit=10" \
  -H "Authorization: Bearer sk_seedancee2_live_xxxxx"

List Models

GET /v1/models

Example response:

{
  "data": [
    {
      "id": "pro",
      "status": "active",
      "resolutions": ["480p", "720p", "1080p"],
      "duration": {
        "min": 4,
        "max": 15
      }
    }
  ]
}

Usage Summary

GET /v1/usage

Supported query parameters:

Field	Type	Description
`start`	string	Optional ISO 8601 start datetime filter
`end`	string	Optional ISO 8601 end datetime filter
`limit`	integer	Max daily timeline rows to return. Default 30, max 365

Example:

curl "https://api.SeeDanceE2.ai/v1/usage?start=2026-07-01T00:00:00.000Z&end=2026-07-31T23:59:59.999Z&limit=7" \
  -H "Authorization: Bearer sk_seedancee2_live_xxxxx"

Example response:

{
  "data": {
    "range": {
      "start": "2026-07-01T00:00:00.000Z",
      "end": "2026-07-31T23:59:59.999Z"
    },
    "summary": {
      "total_jobs": 12,
      "pending_jobs": 1,
      "processing_jobs": 2,
      "completed_jobs": 8,
      "failed_jobs": 1,
      "credits_charged": 280,
      "credits_refunded": 40,
      "net_credits": 240
    },
    "by_model": [
      {
        "model": "mini",
        "total_jobs": 7,
        "pending_jobs": 0,
        "processing_jobs": 1,
        "completed_jobs": 5,
        "failed_jobs": 1,
        "credits_charged": 120,
        "credits_refunded": 20,
        "net_credits": 100
      },
      {
        "model": "pro",
        "total_jobs": 5,
        "pending_jobs": 1,
        "processing_jobs": 1,
        "completed_jobs": 3,
        "failed_jobs": 0,
        "credits_charged": 160,
        "credits_refunded": 20,
        "net_credits": 140
      }
    ],
    "timeline": [
      {
        "date": "2026-07-01",
        "total_jobs": 4,
        "pending_jobs": 0,
        "processing_jobs": 1,
        "completed_jobs": 3,
        "failed_jobs": 0,
        "credits_charged": 100,
        "credits_refunded": 0,
        "net_credits": 100
      }
    ]
  }
}

Webhooks

If callback_url is provided, SeeDanceE2 sends a POST request when the task reaches completed or failed.

Recommended headers to inspect:

X-Seedancee2-Event: video.generation.completed
X-Seedancee2-Signature: t=1770000000,v1=...

If webhook signing is enabled for your account, the body is signed as:

payload_to_sign = "{timestamp}.{raw_json_body}"
signature = hex(hmac_sha256(webhook_secret, payload_to_sign))

Completed webhook payload:

{
  "data": {
    "id": "vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
    "status": "completed",
    "model": "pro",
    "created_at": "2026-06-29T10:00:00.000Z",
    "completed_at": "2026-06-29T10:06:20.000Z",
    "credits_charged": 100,
    "output": {
      "video_url": "https://media.SeeDanceE2.ai/videos/2026/06/29/vid_01jz8b7q4x8k2p9w3m5c6n7r8s.mp4",
      "thumbnail_url": "https://media.SeeDanceE2.ai/video-posters/2026/06/29/vid_01jz8b7q4x8k2p9w3m5c6n7r8s.jpg"
    },
    "metadata": {
      "customer_job_id": "job_123",
      "campaign": "launch_week"
    }
  }
}

Failed webhook payload:

{
  "data": {
    "id": "vid_01jz8b7q4x8k2p9w3m5c6n7r8s",
    "status": "failed",
    "model": "pro",
    "created_at": "2026-06-29T10:00:00.000Z",
    "credits_charged": 0,
    "credits_refunded": 100,
    "error": {
      "code": "generation_failed",
      "message": "The video could not be generated. Please try again with a different prompt.",
      "type": "generation_error"
    },
    "metadata": {
      "customer_job_id": "job_123",
      "campaign": "launch_week"
    }
  }
}

Input Guidelines

Images:

Requirement	Value
Formats	JPEG, PNG, WebP, BMP, TIFF, GIF
Count	Up to 9 reference images
`first_and_last_frames`	Exactly 2 images

Videos:

Requirement	Value
`video_url`	1
`video_urls`	Up to 3

Audio:

Requirement	Value
`audio_urls`	Up to 3

Errors

Error format:

{
  "error": {
    "code": "insufficient_credits",
    "message": "Your credit balance is too low. Please top up.",
    "type": "billing_error"
  }
}

Current public error codes:

HTTP status	Code	Type	Description
400	`invalid_request`	`invalid_request_error`	Missing or invalid parameters
401	`invalid_api_key`	`authentication_error`	API key is missing, invalid, or disabled
402	`insufficient_credits`	`billing_error`	Credit balance too low
404	`model_not_found`	`invalid_request_error`	Model does not exist or is inactive
404	`task_not_found`	`invalid_request_error`	Task ID does not exist
500	`internal_error`	`api_error`	Unexpected server error

Integration Notes

Use only public model names such as pro, fast, and mini.
Do not build logic against internal provider names, internal task IDs, or internal media hosts.
Store your own metadata so you can join callbacks and polling results to internal business records.
For production use, prefer webhooks for completion and keep polling as a fallback.

SeeDanceE2.ai 기반으로 구축하세요.

One API surface for video generation

Bearer API keys

Submit, poll, webhook

Reserved and charged

Use only `pro`, `fast`, and `mini`

Production quality

Faster iteration

Lower-cost tests

Frequently asked questions

SeeDanceE2 Public Video API

Base URL

Quickstart

Authentication

API keys and model access

Public Models

Credits

Task Lifecycle

Create Video Task

Get Video Task

List Video Tasks

List Models

Usage Summary

Webhooks

Input Guidelines

Errors

Integration Notes