Video Model scene
This page helps you select the MoonNexAI video model by business scene. Actual models, prices, aspect ratio and duration are based on model list API and console configuration.
Common Models
| scene | Recommended Model | Annotations |
|---|---|---|
| NT 720P real-person reference video | seedance-2.0-nt, seedance-2.0-nt-fast, seedance-2.0-nt-mini, seedance-2.5-nt | Settled by actual token usage. Verified with one public image, 4 seconds, 16:9. See NT for duration and strict first-frame ratio rules. |
| Seedance Official Token Usage billing and Video Reference | seedance-2.0-sz, seedance-2.0-sz-fast, seedance-2.0-sz-mini, seedance-2.5-sz | Use references[] in a uniform manner; support pictures, videos and audio references, with resolution, duration and media assets ceilings executed by model page. |
| Default Quick Generate | seedance-2.0-kz-fast | Suitable for first run-through process, high frequency generation and bulk task. |
| Standard image-to-video and independent media assets lines | seedance-2.0-cl-fast, seedance-2.0-cl, seedance-2.0-cl-mini | Fits the URL and this model line Asset:// image-to-video; seedance-2.0-cl-mini supports 480P / 720P, 4-15 seconds; for different media assets lines, see the standalone model page. |
| Seedance 2.5 Multimodule and Authentication media assets | seedance-2.5-cl | Supports 4 - 30 seconds, 480P / 720P / 1080P, Photo/Video/Audience Reference and Asset:// prepared by model; direct error is reported when media assets resolution fails, and does not retreat to the public web URL. |
| CL-B multimodular video compatible with Volcengine-compatible parameters | seedance-2.0-cl-b-fast, seedance-2.0-cl-b-mini | Matches with Volcengine-compatible Seedance 2.0 Fast/ Mini parameters and capabilities; supports official content[] and equivalent models for permitted pictures, videos and audio references. |
| Up to 15 multi-modular reference media assets | seedance-2.0-gl, seedance-2.0-gl-fast, seedance-2.0-gl-mini | Supports 4 to 15 seconds, up to 9 + 3 videos + 3 audio; standard version supports 1080 P, Fast/ Mini supports 480 P/720 P. |
| 720P cross-second reference | moon-2.0-dj-fast | Use references[] to support up to 9 photo references; reference images does not require to be a face map; video reference and independent audio are not supported. |
| 480P / 720P generation billed per second | moon-2.5-gl-a | Text-only and mixed references, 4–30 seconds; up to 30 images, 10 videos and 10 audio files, at most 50 references total. |
| Higher quality or 1080P | seedance-2.0-kz | Fits to higher quality, 1080P and more complete media assets control. |
| 720P mixed-media public URLs | moon-2.0-wc-720p, moon-2.0-wc-b-720p, moon-2.0-wc-b-720p-fast | Use references[]. Supports real people and at most 12 references: 9 images, 3 videos, 3 audio files. Base accepts text-only input and 4–14 seconds; a request for 15 generates 14 seconds. B/B Fast require references and support 4–15 seconds. |
| 720 P IX reference | moon-2.0-wc-c-720p | Supports real people, 4 to 15 seconds target duration, plain text and up to 9 reference images; no video and audio. |
| 720P 4-3-1 Mixed reference | moon-2.0-wc-d-720p | Supports real people, 4 to 15 seconds target duration, plain text or up to 4 + 3 video + 1 audio; audio cannot be submitted alone. |
| 720 P WC-E 4-3-1 Mixed reference | moon-2.0-wc-e-720p | Supports real people, 4 to 15 seconds target duration, plain text or up to 4 + 3 video + 1 audio; audio cannot be submitted alone. |
| 720P WC-F Mixed Reference | moon-2.0-wc-f-720p | Supports real people, 4 to 15 seconds, plain text or photo/video/video reference; up to 9 figures, 3 videos, 3 audio, up to 12 media assets; audio must be matched with pictures. |
| 720P character and position reference, multi-media assets generated in descending order | moon-2.0-ld-a, moon-2.0-ld-b | Two models have the same capacity; 5 to 15 seconds support, up to 9 pictures, 3 videos, 3 audio, up to 15; and reference images plus background music for multiple characters. |
| 720P Quick Multi-media assets | moon-2.0-am-720p-fast | Use references[] for single graphs, pictures + background music and multi-photograms, multi-video, multi-sound public network URL combinations. |
| 720P mixed media billed per second | moon-2.0-il-720p | Supports 4–14 seconds. Use references[] with up to 9 images, 3 videos and 3 audio files, at most 12 total. |
| 720P Photo Reference or Mixed Reference | moon-2.0-me-a-720p, moon-2.0-me-b-720p, moon-2.0-me-a-fast | A / A Fast up to 9 pictures; B up to 9 + 3 videos + 3 audio, up to 15; support 5 / 10 / 15 seconds. |
| Use sub-utility text-to-video | happyhorse-1.0 | Support 720P / 1080P, 5/10/15 seconds, first frame and reference images; do not support last frame and reference video. |
| first and last frames and reference video | wan2.7 | Support 720P / 1080P, 5/10/15 seconds, first frame, last frame, reference images and reference video. |
| Sora style task | sora-v3-fast, sora-v3-pro | Fits to text-to-video, reference images and Remix scenarios. |
| Grok Video | grok-video-1.5, grok-image-video | grok-video-1.5 Current API behavior requires single reference images; grok-image-video supports up to 5 reference images for up to 15 seconds. |
| Qwen Video | Model list API | Video generation and video extension appropriate for the corresponding model. |
media assets Mode
| Input Method | Recommended use |
|---|---|
references[] | Recommended for new mixed-media integrations; describes image, video, audio and music references in one format. |
images[] | Compatible arrays broken down by type of picture: multiple pictures, first frame, last frame, reference images. |
videos[] | reference video or video to be extended. |
audios[] | Use when the model supports separate audio references. |
content[] | Use when mixing text, pictures, videos, audio or reference images binding audio. |
Asset://asset_xxx | The same media assets is repeated over and the target model explicitly supports the media assets reference. |
Example: Moon-2 multi-media assets video
json
{
"model": "moon-2.0-wc-720p",
"prompt": "根据参考图生成一段动作自然的视频。",
"references": [
{
"media_type": "image",
"role": "reference_image",
"url": "https://example.com/reference.png",
"alias": "hero"
}
],
"duration": 14,
"ratio": "9:16"
}Example: High-quality first frame video
json
{
"model": "seedance-2.0-kz",
"prompt": "保持人物外观一致，动作自然流畅，镜头稳定。",
"images": [
{
"url": "https://example.com/first-frame.png",
"role": "first_frame"
}
],
"duration": 10,
"resolution": "1080P",
"ratio": "9:16"
}Relevant Pages
| Page | Annotations |
|---|---|
| Video Generation Guide | Generic video task, query and media assets description. |
| Seedance Video Guide | The selection of the current open model and the media assets proposal. |
| Seedance-2 Video Overview | Seedance model page entry. |
| Moon-2 Video Overview | Moon-2 model page entry. |
| Video Model Matrix | Model capacity matrix. |
| HappyHorse / Wan Video Model | happyhorse-1.0 and wan2.7 parameters, media assets boundary and examples. |
| Grok Video | grok-video-1.5 and grok-image-video parameters, reference images boundary and examples. |