seedance-2.0-cl-b-fast / seedance-2.0-cl-b-mini

seedance-2.0-cl-b-fast and seedance-2.0-cl-b-mini are separate Seedance-2 CL-B public model lines for MoonNexAI. Both use independent public model names and capabilities to verify that CL-B models are not replaced by seedance-2.0-cl, seedance-2.0-cl-fast or seedance-2.0-cl-mini.

Entry and Model Name

Item	Annotations
Recommended entrance	POST /v1/videos
Official shape compatible entrance	POST /api/v3/contents/generations/tasks
Query endpoint	GET /v1/videos/{task_id}, GET /v1/tasks/{task_id} or the corresponding compatible task-query path
Quick Model	seedance-2.0-cl-b-fast
Lightweight model	seedance-2.0-cl-b-mini

The compatibility entry retains the Volcengine official request shape, but model must fill out the name of the MoonNexAI public model in the above table. Do not fill in the client request private model name or service-end identifier.

Compatibility of official parameters

Both CL-B models follow the Volcengine Seedance 2.0 parameter contract. Clients can retain the official request fields and nested structure, but must set model to the appropriate MoonNexAI public model ID.

Capacity	seedance-2.0-cl-b-fast	seedance-2.0-cl-b-mini
Official request shape	Support for official field, content[], ratio, resolution, duration, generate_audio, watermark, seed, callback_url	Same Left
text-to-video	Supported by official modelling capacity	Supported by official modelling capacity
Picture reference	Supported by official modelling capacity	Supported by official modelling capacity
Video reference	Supported by official modelling capacity	Supported by official modelling capacity
Audio Reference	Supported by official modelling capacity	Supported by official modelling capacity
aspect ratio, resolution, duration	Consistent with official Fast capacity matrix	Consistent with the official Mini capacity matrix
Audio Generation Switches	generate_audio corresponds to official semantics	generate_audio corresponds to official semantics

The availability of parameters, the range of values to be taken and the media assets quantitative limit are based on the official capacity matrix of the selected model; the Fast type limits are not applied to Mini, or vice versa.

Subbar for request Parameters

MoonNexAI Unified Universal Video Parameters

The entry is POST /v1/videos, using prompt, references[], aspect_ratio and size.

field	Type	Annotations
model	string	seedance-2.0-cl-b-fast or seedance-2.0-cl-b-mini.
prompt	string	Video description, with @alias references to media assets.
references[]	array	General image, video and audio media assets arrays.
duration	number	By the selected official Fast/ Mini model.
aspect_ratio	string	Video aspect ratio.
size	string	Output resolution.
generate_audio	boolean	Audio generation switches.
watermark	boolean	Watermark switch.
callback_url	string	Optional callback address.

Volcengine-compatible compatible parameters

The compatible endpoint is POST /api/v3/contents/generations/tasks, using content[], ratio and resolution. Do not submit both reference formats in the same request.

field	Type	Annotations
model	string	The name of the MoonNexAI public model is still being filled in for CL-B.
content[]	array	Official text, pictures, videos and audio entries.
duration	number	Compatible duration field.
ratio	string	Compatible aspect-ratio field.
resolution	string	Compatible resolution field.
generate_audio	boolean	Official audio switch.
watermark	boolean	Official watermark switch.
seed	integer	Official random seeds.
callback_url	string	Official callback address.

Officially request

To maintain the formal prompt semantics of media assets, the media entry should be marked to role and the only alias. alias does not contain @, and @Image1, @Video1 or @Audio1 references are used in prompt.

curl https://moonnexai.com/api/v3/contents/generations/tasks \
  -H "Authorization: Bearer <MOONNEXAI_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2.0-cl-b-fast",
    "content": [
      {
        "type": "text",
        "text": "让主体自然移动，镜头平稳推进。"
      },
      {
        "type": "image_url",
        "role": "reference_image",
        "image_url": {
          "url": "https://example.com/reference.png"
        },
        "alias": "Image1"
      },
      {
        "type": "video_url",
        "role": "reference_video",
        "video_url": {
          "url": "https://example.com/motion.mp4"
        },
        "alias": "Video1"
      },
      {
        "type": "audio_url",
        "role": "reference_audio",
        "audio_url": {
          "url": "https://example.com/music.mp3"
        },
        "alias": "Audio1"
      }
    ],
    "duration": 4,
    "ratio": "16:9",
    "resolution": "720P",
    "generate_audio": true,
    "watermark": false
  }'

model is the lightweight model that is selected by changing to seedance-2.0-cl-b-mini. content[] and references[] are two input expressions; the same request is not submitted simultaneously.

Multigraph + MultiVideo + Multi Audio

CL-B accepts multiple references according to the selected Fast or Mini model's capability matrix. The example uses 2 images, 2 videos and 2 audio files. Larger combinations must still satisfy that model's per-type and total limits.

{
  "model": "seedance-2.0-cl-b-fast",
  "content": [
    { "type": "text", "text": "结合 @Image1 和 @Image2 的主体外观，参考 @Video1 与 @Video2 的动作节奏，并将 @Audio1、@Audio2 作为声音参考。" },
    { "type": "image_url", "role": "reference_image", "image_url": { "url": "https://example.com/subject-1.png" }, "alias": "Image1" },
    { "type": "image_url", "role": "reference_image", "image_url": { "url": "https://example.com/subject-2.png" }, "alias": "Image2" },
    { "type": "video_url", "role": "reference_video", "video_url": { "url": "https://example.com/motion-1.mp4" }, "alias": "Video1" },
    { "type": "video_url", "role": "reference_video", "video_url": { "url": "https://example.com/motion-2.mp4" }, "alias": "Video2" },
    { "type": "audio_url", "role": "reference_audio", "audio_url": { "url": "https://example.com/music-1.mp3" }, "alias": "Audio1" },
    { "type": "audio_url", "role": "reference_audio", "audio_url": { "url": "https://example.com/music-2.mp3" }, "alias": "Audio2" }
  ],
  "duration": 4,
  "ratio": "16:9",
  "resolution": "720P",
  "generate_audio": true
}

content[].alias must be the only one in the same request; the same six entries can also be converted to references[], but do not submit content and references simultaneously in a request.

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

{
  "model": "seedance-2.0-cl-b-fast",
  "prompt": "结合 @Image1 和 @Image2 的主体外观，参考 @Video1 与 @Video2 的动作节奏，并将 @Audio1、@Audio2 作为声音参考。",
  "references": [
    { "media_type": "image", "role": "reference_image", "url": "https://example.com/subject-1.png", "alias": "Image1" },
    { "media_type": "image", "role": "reference_image", "url": "https://example.com/subject-2.png", "alias": "Image2" },
    { "media_type": "video", "role": "reference_video", "url": "https://example.com/motion-1.mp4", "alias": "Video1" },
    { "media_type": "video", "role": "reference_video", "url": "https://example.com/motion-2.mp4", "alias": "Video2" },
    { "media_type": "audio", "role": "reference_audio", "url": "https://example.com/music-1.mp3", "alias": "Audio1" },
    { "media_type": "audio", "role": "reference_audio", "url": "https://example.com/music-2.mp3", "alias": "Audio2" }
  ],
  "duration": 4,
  "aspect_ratio": "16:9",
  "size": "720P",
  "generate_audio": true
}

Uniform access

When you need to use MoonNexAI to unify video API, you can submit the same media syntax:

{
  "model": "seedance-2.0-cl-b-mini",
  "prompt": "让主体自然移动，镜头平稳推进。",
  "references": [
    {
      "media_type": "image",
      "role": "reference_image",
      "url": "https://example.com/reference.png",
      "alias": "Image1"
    },
    {
      "media_type": "video",
      "role": "reference_video",
      "url": "https://example.com/motion.mp4",
      "alias": "Video1"
    },
    {
      "media_type": "audio",
      "role": "reference_audio",
      "url": "https://example.com/music.mp3",
      "alias": "Audio1"
    }
  ],
  "duration": 4,
  "aspect_ratio": "16:9",
  "resolution": "720P",
  "generate_audio": true
}

Pictures, videos and audio URLs must be a stable public media address that the service can access anonymously and Content-Type corresponds to the file type. Uploading or preparing media assets does not automatically attach to subsequent task; each generation must be visible to the URL or to the media assets reference that the model clearly supports.

billing and Results

CL-B settles by Seedance-2 token usage. Check the console, model-list API and actual billing records for prices and available groups. Query status after creation and read url, video_url or output.url only for succeeded.

Relevant Pages

• Seedance-2 Video Overview
• Video Model Matrix
• Create Video Job
• Query Video Job
• Lookdance-2.0-cl/ CL series