Seedance-2.0 Original Series

This series consists of three independent public models that are unified by POST /v1/videos to create task:

This page uses the Volcanic Open Model Name directly. A specific site or account has been activated for a model, which is based on the current console Model List and the price page.

Models	Positioning	resolution	duration
`doubao-seedance-2-0-260128`	Standard version for high quality and `1080P` scenes	`480P`, `720P`, `1080P`	`4` - `15` seconds
`doubao-seedance-2-0-fast-260128`	Fast version, suitable for high frequency generation	`480P`, `720P`	`4` - `15` seconds
`doubao-seedance-2-0-mini-260615`	Light version, suitable for mass generation	`480P`, `720P`	`4` - `15` seconds

All three models support text-to-video and image, video and audio references. Standard and Fast cover broader multimodal creation. Mini uses the same public request structure but outputs only 480P or 720P.

Subbar for request Parameters

MoonNexAI Unified Common Parameters	Volcengine-compatible compatible parameters
`model`	`model` (Put in the public model of this page ID)
`prompt`	`text` Entry for `content[]`
`references[]`	`image_url`, `video_url`, `audio_url` entries in `content[]`
`aspect_ratio`, `size`	`ratio`, `resolution`
`duration`, `generate_audio`, `watermark`, `seed`, `callback_url`	Same official field, visible zero and `false` reserved

MoonNexAI Unified Universal Video Parameters

field	Annotations
`model`	One of three original open models.
`prompt`	Video description, with `@alias` quote media assets.
`references[]`	Picture, video and audio uniform media assets arrays, entries `media_type`, `role`, `url`, `alias`.
`aspect_ratio` / `size`	Common aspect-ratio and resolution fields, mapped to model parameters by the server.
`duration`	`4` - `15` sec.
`generate_audio`, `watermark`, `seed`, `callback_url`	General control field, verified by model capability.

Volcengine-compatible compatible parameters

field	Annotations
`model`	Fill in one of three MoonNexAI public models IDs on this page.
`content[]`	Official multi-modular arrays, with `text`, `image_url`, `video_url` or `audio_url`, media entries `role` and only `alias`.
`ratio` / `resolution`	Compatible aspect-ratio and resolution fields.
`duration`, `generate_audio`, `watermark`, `seed`, `callback_url`	Same as Volcengine-compatible.

Basic request (MoonNexAI Unified Universal Parameters)

bash

curl https://moonnexai.com/v1/videos \
  -H "Authorization: Bearer <MOONNEXAI_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedance-2-0-fast-260128",
    "prompt": "清晨海边，镜头平稳向前推进，浪花自然拍打礁石。",
    "duration": 5,
    "resolution": "720p",
    "ratio": "16:9",
    "generate_audio": false,
    "watermark": false
  }'

Base request (Volcengine-compatible compatible parameters)

bash

curl https://moonnexai.com/api/v3/contents/generations/tasks \
  -H "Authorization: Bearer <MOONNEXAI_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedance-2-0-fast-260128",
    "content": [
      { "type": "text", "text": "清晨海边，镜头平稳向前推进，浪花自然拍打礁石。" }
    ],
    "duration": 5,
    "resolution": "720p",
    "ratio": "16:9",
    "generate_audio": false,
    "watermark": false
  }'

Saves task_id or id after creating successfully, and then calls the GET /v1/videos/{task_id} query. Only the final video address is read when task is succeeded.

Use reference media assets

New access is recommended for references[], which provides each media assets visible medium type and role:

json

{
  "model": "doubao-seedance-2-0-mini-260615",
  "prompt": "让 @product 在摄影棚转台上缓慢旋转，保持产品外观一致。",
  "references": [
    {
      "media_type": "image",
      "role": "reference_image",
      "url": "https://example.com/product.png",
      "alias": "product"
    }
  ],
  "duration": 5,
  "size": "720p",
  "aspect_ratio": "16:9",
  "generate_audio": false
}

Video and audio references should meet the duration, format and file size requirements of the target model.

Multigraph + MultiVideo + Multi Audio (MoonNexAI Unified Generic Parameters)

The three primary models use the same multimodular request structure. The following example uses 2 pictures, 2 videos and 2 segments of audio; please adjust the number and resolution to the actual ability of the model list to return.

json

{
  "model": "doubao-seedance-2-0-fast-260128",
  "prompt": "让 @subject1、@subject2 保持外观一致，参考 @clip1、@clip2 的动作，并将 @sound1、@sound2 融合为背景声音。",
  "references": [
    { "media_type": "image", "role": "reference_image", "url": "https://example.com/subject-1.png", "alias": "subject1" },
    { "media_type": "image", "role": "reference_image", "url": "https://example.com/subject-2.png", "alias": "subject2" },
    { "media_type": "video", "role": "reference_video", "url": "https://example.com/clip-1.mp4", "alias": "clip1" },
    { "media_type": "video", "role": "reference_video", "url": "https://example.com/clip-2.mp4", "alias": "clip2" },
    { "media_type": "audio", "role": "reference_audio", "url": "https://example.com/sound-1.mp3", "alias": "sound1" },
    { "media_type": "audio", "role": "reference_audio", "url": "https://example.com/sound-2.mp3", "alias": "sound2" }
  ],
  "duration": 5,
  "size": "720p",
  "aspect_ratio": "16:9",
  "generate_audio": true
}

Audio references must be submitted with pictures or videos; Asset://asset_xxx must also be prepared for generation_ready=true using the actual primary model.

Multigraph + MultiVideo + Multi Audio (Volcengine-compatible compatible parameters)

Group 2 + 2 Video + 2 Audio to official content[] shape:

json

{
  "model": "doubao-seedance-2-0-fast-260128",
  "content": [
    { "type": "text", "text": "让 @subject1、@subject2 保持外观一致，参考 @clip1、@clip2 的动作，并融合 @sound1、@sound2。" },
    { "type": "image_url", "role": "reference_image", "image_url": { "url": "https://example.com/subject-1.png" }, "alias": "subject1" },
    { "type": "image_url", "role": "reference_image", "image_url": { "url": "https://example.com/subject-2.png" }, "alias": "subject2" },
    { "type": "video_url", "role": "reference_video", "video_url": { "url": "https://example.com/clip-1.mp4" }, "alias": "clip1" },
    { "type": "video_url", "role": "reference_video", "video_url": { "url": "https://example.com/clip-2.mp4" }, "alias": "clip2" },
    { "type": "audio_url", "role": "reference_audio", "audio_url": { "url": "https://example.com/sound-1.mp3" }, "alias": "sound1" },
    { "type": "audio_url", "role": "reference_audio", "audio_url": { "url": "https://example.com/sound-2.mp3" }, "alias": "sound2" }
  ],
  "duration": 5,
  "resolution": "720p",
  "ratio": "16:9",
  "generate_audio": true
}

Create reusable media assets

Normal pictures, videos, and audio references media assets can be directly called to POST /v1/assets/uploads for creation, without live authentication. mode=asset is recommended and the model to be used when actually generated is completed:

bash

curl https://moonnexai.com/v1/assets/uploads \
  -H "Authorization: Bearer <MOONNEXAI_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedance-2-0-260128",
    "mode": "asset",
    "type": "image",
    "purpose": "reference_image",
    "role": "reference_image",
    "url": "https://example.com/reference.png"
  }'

Ask for media assets details after creation. Only if the following conditions are met:

text

generation_ready = true
preparation_status = ready

Explicitly pass the returned Asset://asset_xxx when creating the video task:

json

{
  "model": "doubao-seedance-2-0-260128",
  "prompt": "保持 @hero 的外观和服装一致，让人物自然转身。",
  "references": [
    {
      "media_type": "image",
      "role": "reference_image",
      "url": "Asset://asset_xxx",
      "alias": "hero"
    }
  ],
  "duration": 5,
  "resolution": "720p",
  "ratio": "16:9"
}

Common reference media assets is not the same process as private reality. Common media media assets does not require live authentication; private property that requires a specific identity, long-term re-image of a person, is restricted, requires a separate authorization and identity, and cannot be replaced by common media assets.

media assets preparation results are separated by model. Create or confirm media assets by the primary model that generates request, which is actually used, and do not mix media assets directly with other models.

List of Parameters

MoonNexAI Unified Universal field

field	Annotations
`model`	Fill in one of the three public models of the page.
`prompt`	Video description, with `@alias` quote media assets.
`references[]`	Photo, video and audio unified reference arrays.
`aspect_ratio` / `size`	Universal aspect ratio and output size field.
`duration`	`4` - `15` sec.
`generate_audio` / `watermark`	Audio and watermark switches.
`seed` / `camera_fixed`	Optional control field.

Volcengine-compatible compatible field

field	Annotations
`model`	Fill in the public model ID of this page.
`content[]`	Official multi-modular input arrays.
`ratio` / `resolution`	Compatible aspect ratio and output resolution.
`duration`	Compatible duration field.
`generate_audio` / `watermark` / `seed`	Official control of field.

Billing

All three models settle by output tokens, distinguishing requests with and without video input. Check the console, pricing API and actual settlement records for current prices by resolution, version and input type. Documentation does not fix monetary amounts.

Conversation APIs

Embeddings

Audio

Models

GPT Image

Gemini Nano

Ideogram Image

Midjourney Task

Seedance-2 Video

Moon-2 Video

Seedance-2.0 Original Series

Subbar for request Parameters

MoonNexAI Unified Universal Video Parameters

Volcengine-compatible compatible parameters

Basic request (MoonNexAI Unified Universal Parameters)

Base request (Volcengine-compatible compatible parameters)

Use reference media assets

Multigraph + MultiVideo + Multi Audio (MoonNexAI Unified Generic Parameters)

Multigraph + MultiVideo + Multi Audio (Volcengine-compatible compatible parameters)

Create reusable media assets

List of Parameters

MoonNexAI Unified Universal field

Volcengine-compatible compatible field

Billing

Relevant Pages

Ideogram Image

Midjourney Task

Seedance-2 Video

Moon-2 Video

Seedance-2.0 Original Series ​

Subbar for request Parameters ​

MoonNexAI Unified Universal Video Parameters ​

Volcengine-compatible compatible parameters ​

Basic request (MoonNexAI Unified Universal Parameters) ​

Base request (Volcengine-compatible compatible parameters) ​

Use reference media assets ​

Multigraph + MultiVideo + Multi Audio (MoonNexAI Unified Generic Parameters) ​

Multigraph + MultiVideo + Multi Audio (Volcengine-compatible compatible parameters) ​

Create reusable media assets ​

List of Parameters ​

MoonNexAI Unified Universal field ​

Volcengine-compatible compatible field ​

Billing ​

Relevant Pages ​

Seedance-2.0 Original Series

Subbar for request Parameters

MoonNexAI Unified Universal Video Parameters

Volcengine-compatible compatible parameters

Basic request (MoonNexAI Unified Universal Parameters)

Base request (Volcengine-compatible compatible parameters)

Use reference media assets

Multigraph + MultiVideo + Multi Audio (MoonNexAI Unified Generic Parameters)

Multigraph + MultiVideo + Multi Audio (Volcengine-compatible compatible parameters)

Create reusable media assets

List of Parameters

MoonNexAI Unified Universal field

Volcengine-compatible compatible field

Billing

Relevant Pages