Seedance Video Guide

Seedance-2 provides eight separate series of SG, NT, SZ, birth, KZ, CL, CL-B and GL. New access first determines the series and then selects it from the public model names listed on this series page; the parameters of the different series and the media assets preparation process are not mixed.

The SG family includes seedance-2.0-sg-mini, seedance-2.0-sg, seedance-2.0-sg-fast and seedance-2.5-sg, supporting common and compatible parameters. All four have verified single-image, 4-second generation. Standard supports 480P / 720P / 1080P / 4K, Fast/Mini 480P / 720P, and 2.5 480P / 720P / 1080P. Billing uses output tokens.

The four public SZ models, seedance-2.5-cl and both public CL-B models accept Volcengine-compatible content[], ratio, resolution, duration, generate_audio and related fields. To retain the Volcengine-compatible request path, use /api/v3/contents/generations/tasks, but model still fills out the MoonNexAI public model ID, e. g. seedance-2.0-sz, seedance-2.5-sz, seedance-2.5-cl or seedance-2.0-cl-b-fast. Both entry points use MoonNexAI API Key, model still press this page Public model list.

seedance-2.5-cl recommends the use of /v1/videos and references[], which also support the official shape content[]. field correspondence, full examples and authentication of media assets rules are available in official compatibility-2.5-cl parameters; this does not mean that all official advanced parameters are available, and the specific capabilities are based on model pages.

The parameters and capabilities of CL-B are aligned to the official Volcengine Seedance 2.0 Fast/ Mini compact. Media entries suggest that the only alias is provided in a visible format; field takes values, media assets numbers and model boundaries. Please follow CL-B Series Page and the selected official model capability matrix.

Quick Select

Requirements	Recommended Model	Annotations
NT 720P True reference images Settlement with Token	`seedance-2.0-nt`, `seedance-2.0-nt-fast`, `seedance-2.0-nt-mini`, `seedance-2.5-nt`	Four models have been validated for single grid reference images, 4 seconds and 16:9; Mini is only 4 seconds, 2.5 strict first frame maintains input aspect ratio. See NT Series.
Official Token Multimodular Video and Video Reference with billing Volume	`seedance-2.0-sz`, `seedance-2.0-sz-fast`, `seedance-2.0-sz-mini`, `seedance-2.5-sz`	The SZ series uses `references[]`; the 2.0 series supports `4` - `15` seconds, and 2.5 supports `4` - `30` seconds. When including video references, enter billing dimensions on official video.
Volcanic open primary model	`doubao-seedance-2-0-260128`, `doubao-seedance-2-0-fast-260128`, `doubao-seedance-2-0-mini-260615`	The three models of the original series, Fast, Mini, with specific parameters and media assets processes available on a stand-alone page.
Default Quick Generate	`seedance-2.0-kz-fast`	Fits to HF, batch and default quick-out.
Public URL or this model line `Asset://` image-to-video/ Video Reference	`seedance-2.0-cl-fast` or `seedance-2.0-cl`	Standard compatible circuits suitable for single graph image-to-video and reference video task; video reference suggests that this model line media assets ID, media assets ID be created first, not mixed with the walrus walrus walrus of walrus.
Seedance 2.5 Multimodule and Authentication media assets	`seedance-2.5-cl`	Supports 4 - 30 seconds, 480P / 720P / 1080P, up to 50 mixed reference media assets; `Asset://` shall be prepared and maintained in this model and resolve failure to report direct error.
Light 480P / 720P CL Workstream	`seedance-2.0-cl-mini`	Parameters are consistent with the CL series, supporting text-to-video, image-to-video, video references and this model line `Asset://` media assets quotes, duration `4` - `15` seconds.
Volcengine-compatible-shaped CL-B mix media assets	`seedance-2.0-cl-b-fast`, `seedance-2.0-cl-b-mini`	Use `/api/v3/contents/generations/tasks` and official `content[]`; parameters, aspect ratio, resolution, duration and media assets borders are consistent with selected official Fast/ Mini contracts.
Mixed-media references	`seedance-2.0-gl`, `seedance-2.0-gl-fast`, `seedance-2.0-gl-mini`	Up to 9 images + 3 videos + 3 audio files, at most 15 total; `4`–`15` seconds. Standard supports up to 1080P; Fast/Mini support 480P / 720P.
Higher quality or 1080P	`seedance-2.0-kz`	Fits to a higher quality and 1080 P scene.

Two request formats

NT also accepts /api/v3/contents/generations/tasks with compatible content[], ratio, resolution, duration, generate_audio and watermark, using NT public model IDs and a MoonNexAI API key. See NT compatible parameters for the single-image example, durations and authenticated-asset limits. Poll for results.

MoonNexAI provides two sets of equivalents of Seidance request expression. Each time request selects one set, do not both references[] and content[], and do not repeat prompt and content[].text.

MoonNexAI Unified Universal Video Parameters

Common access POST /v1/videos:

Parameters	Annotations
`model`	MoonNexAI Public Model Name.
`prompt`	Video description, with `@alias` references to media assets.
`references[]`	Common media array with `media_type`, `role`, `url` and `alias` on each entry.
`duration`	Target duration.
`aspect_ratio`	Video aspect ratio.
`size`	Output size.
`generate_audio`, `watermark`, `seed`, `callback_url`	Universal control field.

Volcengine-compatible compatible parameters

The official shape parameters can be submitted to POST /v1/videos, or can be used to use compatible access POST /api/v3/contents/generations/tasks:

Parameters	Annotations
`model`	The name of the MoonNexAI public model is still being filled in.
`content[]`	An array of official text and media entries.
`duration`	Compatible duration field.
`ratio`	Compatible aspect-ratio field.
`resolution`	Compatible resolution field.
`generate_audio`, `watermark`, `seed`, `callback_url`	Official control of field.

Universal media assets Reference

Prefer references[] for new mixed-media integrations. Specify image, video, audio or music through media_type, purpose through role, and an optional prompt reference name through alias. images[], videos[], audios[] and content[] remain compatible formats.

json

{
  "model": "seedance-2.0-kz-fast",
  "prompt": "让人物自然转身，镜头稳定推进。",
  "references": [
    {
      "media_type": "image",
      "role": "reference_image",
      "url": "https://example.com/reference.png",
      "alias": "hero"
    }
  ],
  "duration": 14,
  "ratio": "16:9"
}

When multi-media assets is combined, multiple pictures, videos and audio segments can be inserted in the same references[]. Below are two + 2 videos + 2 audio examples that are used across the series; availability and individual/total upper limit, whichever is the details page of the selected model:

json

{
  "model": "seedance-2.0-gl",
  "prompt": "保持 @image1、@image2 的主体特征，参考 @video1、@video2 的动作，并融合 @audio1、@audio2。",
  "references": [
    { "media_type": "image", "role": "reference_image", "url": "https://example.com/image-1.png", "alias": "image1" },
    { "media_type": "image", "role": "reference_image", "url": "https://example.com/image-2.png", "alias": "image2" },
    { "media_type": "video", "role": "reference_video", "url": "https://example.com/video-1.mp4", "alias": "video1" },
    { "media_type": "video", "role": "reference_video", "url": "https://example.com/video-2.mp4", "alias": "video2" },
    { "media_type": "audio", "role": "reference_audio", "url": "https://example.com/audio-1.mp3", "alias": "audio1" },
    { "media_type": "audio", "role": "reference_audio", "url": "https://example.com/audio-2.mp3", "alias": "audio2" }
  ],
  "duration": 8,
  "ratio": "16:9",
  "resolution": "720P",
  "generate_audio": true
}

All Volcengine-compatible Seedance series can follow this multi-graph, multi-video, multi-sound model; the number of different models, duration, resolution and role details are verified by the corresponding model ' s configuration capabilities, and the client does not need to rewrite the media assets array structure.

Corresponding Volcengine-compatible compatible writings (field synonyms, except for request entry and media assets packagings):

json

{
  "model": "seedance-2.0-gl",
  "content": [
    { "type": "text", "text": "保持 @image1、@image2 的主体特征，参考 @video1、@video2 的动作，并融合 @audio1、@audio2。" },
    { "type": "image_url", "role": "reference_image", "image_url": { "url": "https://example.com/image-1.png" }, "alias": "image1" },
    { "type": "image_url", "role": "reference_image", "image_url": { "url": "https://example.com/image-2.png" }, "alias": "image2" },
    { "type": "video_url", "role": "reference_video", "video_url": { "url": "https://example.com/video-1.mp4" }, "alias": "video1" },
    { "type": "video_url", "role": "reference_video", "video_url": { "url": "https://example.com/video-2.mp4" }, "alias": "video2" },
    { "type": "audio_url", "role": "reference_audio", "audio_url": { "url": "https://example.com/audio-1.mp3" }, "alias": "audio1" },
    { "type": "audio_url", "role": "reference_audio", "audio_url": { "url": "https://example.com/audio-2.mp3" }, "alias": "audio2" }
  ],
  "duration": 8,
  "ratio": "16:9",
  "resolution": "720P",
  "generate_audio": true
}

media assets Upload

URLs are distributed directly in video request when a stable public network URL already exists.
When combining local files, create temporary tests URLs with /v1/assets/uploads mode=oss; use the stable public network URLs that the customer has stored for the official business.
Asset://asset_xxx is only used when the model page explicitly supports the media assets reference.
For seedance-2.0-cl / seedance-2.0-cl-fast/ seedance-2.0-cl-mini, the photo reference can be sent directly to the URL of the public network; the video reference proposal first creates Asset://asset_xxx of type=video, role=reference_video, with /v1/assets/uploads and then passes to video_url or references[].url.
For seedance-2.0-gl / seedance-2.0-gl-fast/seedance-2.0-gl-mini, the client 's own stable public network URL priority; Asset://asset_xxx is created by the target GL model when media assets is required, and audio references must be submitted with pictures or videos.

bash

curl https://moonnexai.com/v1/assets/uploads \
  -H "Authorization: Bearer <MOONNEXAI_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2.0-cl",
    "mode": "asset",
    "type": "video",
    "purpose": "reference_video",
    "role": "reference_video",
    "url": "https://example.com/reference.mp4"
  }'

json

{
  "model": "seedance-2.0-cl",
  "prompt": "保留参考视频中的动作，把上衣图案替换为参考图片风格。",
  "video_url": "Asset://asset_video_xxx",
  "image_url": "https://example.com/reference-shirt.png",
  "duration": 5
}

Conversation APIs

Embeddings

Audio

Models

GPT Image

Gemini Nano

Ideogram Image

Midjourney Task

Seedance-2 Video

Moon-2 Video

Seedance Video Guide

Quick Select

Two request formats

MoonNexAI Unified Universal Video Parameters

Volcengine-compatible compatible parameters

Universal media assets Reference

media assets Upload

Relevant Pages

Ideogram Image

Midjourney Task

Seedance-2 Video

Moon-2 Video

Seedance Video Guide ​

Quick Select ​

Two request formats ​

MoonNexAI Unified Universal Video Parameters ​

Volcengine-compatible compatible parameters ​

Universal media assets Reference ​

media assets Upload ​

Relevant Pages ​

Seedance Video Guide

Quick Select

Two request formats

MoonNexAI Unified Universal Video Parameters

Volcengine-compatible compatible parameters

Universal media assets Reference

media assets Upload

Relevant Pages