Generate Video from Text or References · Alibaba

Describe a clip or provide eligible reference media, a document or a public webpage to generate video with standard Wan3.0. Choose output duration and resolution, with or without generated audio.

Your AI

Use with your AI

For tasks with several steps, continue in the AI you already use.

CodexClaude CodeCursorWindsurfCline

View instructions

Use Toolmu generate-multimodal-video-alibaba (standard Wan3.0 Video, ID e8e66a99-ca39-4c23-b3c5-f08b4f52af75) to reinterpret my short clip. Reuse the existing connection; if setup is needed, follow https://toolmu.com/SKILL.md within my existing execution authorization and budget. First ask me for a real public HTTPS URL of my eligible five-second MP4; do not invent a URL or use a demo clip. Put one object in body.media with type=reference_video and url equal to the supplied link. Set body.prompt to: Reimagine Video 1 as clay animation, using its scene and movement as references. Set body.duration=5, body.resolution=720P, body.ratio=adaptive, body.audio=false and body.prompt_extend=true. The result should be silent; audio=false does not preserve the original soundtrack. This example has five input-video seconds and five output seconds, totaling ten within the 30-second limit; video and audio reference inputs combined must also stay within 15 seconds. Perform only this authorized request. Show the actual successful finished MP4 and available download link, and remind me to inspect the result and save it promptly. Submission alone does not establish completion; do not claim a fixed turnaround or start additional generation automatically.

Task:
Reimagine [my five-second video] in [a clay-animation style]. I will share [a link that lets anyone watch the video itself without signing in]. Please make [a five-second version without sound], then show me the finished video and a link to save it.

Paste into your connected AI's conversation.

Connection guide

Up to 7.2 credits / call

The final charge is based on confirmed usage. Unused reserved credits return to your available balance.

Alibaba

Generate Video from Text or References · Alibaba

Describe a clip or provide eligible reference media, a document or a public webpage to generate video with standard Wan3.0. Choose output duration and resolution, with or without generated audio.

You receive

After successful processing, receive one downloadable H.264 MP4. For a video-to-video task, supply a real public HTTPS link to your five-second MP4 and request a five-second clay-animation interpretation, then review the actual result.

Start with your task

Reimagine [my five-second video] in [a clay-animation style]. I will share [a link that lets anyone watch the video itself without signing in]. Please make [a five-second version without sound], then show me the finished video and a link to save it.

What to know

Input combinations and reference-clip limits apply, and reference video plus output duration cannot exceed 30 seconds. A submitted request is not a finished video. Motion, subject details, lettering and audio fidelity are not guaranteed; processing time varies and new download links typically last about 24 hours.

Inputs & outputs

Inspect the current inputs, availability, and maximum charge before starting.

Input schema

Input schema
Field / Path	Type	Requirement	Meaning
body`body`	`object`	Optional
Random seed`body.seed`	`integer`	Optional	Optional integer from 0 to 2147483647. Helps reuse a random starting value, without guaranteeing identical motion, subjects or videos. Minimum `0` Maximum `2147483647`
Generated audio track`body.audio`	`boolean`	Optional	Optional boolean, default true. true enables an output audio track; false requests silent video. It does not promise that the original source soundtrack will be preserved. Default `true`
Input materials`body.media`	`array`	Optional	Supply an eligible nonempty array if there is no prompt; otherwise optional. Every entry needs type and url. Image, video and audio references are ordered separately by type. Each reference clip is 1–15 seconds, and this Tool limits combined video-and-audio input duration to 15 seconds. For text-only generation, omit media.
[]`body.media[]`	`object`	—	Required fields `type, url`
Public media URL and file limits`body.media[].url`	`string`	—	Required nonempty public HTTPS URL; no HTTP, base64 or local file path. Images: JPG/JPEG, PNG without alpha, BMP or WEBP, ≤20 MB, sides 240–8000 pixels, aspect 1:8–8:1. Videos: MP4/MOV, ≤100 MB, 1–15 seconds, ≥16 fps, sides 240–4096 pixels, aspect 1:8–8:1. Audio: WAV/MP3, ≤15 MB, 1–15 seconds. Files: docx/xlsx/pptx/pdf/txt/md, ≤100 MB and up to 50 pages where applicable. A link must be a public webpage requiring no login. Minimum length `1`
Material type and combinations`body.media[].type`	`string`	—	Required: first_frame, last_frame, reference_image, reference_video, reference_audio, file or link. Use one first frame alone or with one last frame, never last-only or mixed with references/files/links. References may combine: up to 10 images, 5 videos and 5 audio clips. They may also use one file or one link; file and link are mutually exclusive, with at most one of each. Options `first_frame, last_frame, reference_image, reference_video, reference_audio, file, link`
Output aspect ratio`body.ratio`	`string`	Optional	Optional adaptive (default), 16:9, 9:16, 1:1, 4:3 or 3:4. adaptive lets the model choose a suitable ratio from the input and intent; 21:9 is not an exposed choice. Default `adaptive` Options `adaptive, 16:9, 9:16, 1:1, 4:3, 3:4`
Video or editing description`body.prompt`	`string`	Optional	Nonempty text up to 20000 characters when supplied. At least prompt or eligible media is needed. Describe a new clip or the edit; refer to reference assets as Image 1, Video 1 and Audio 1, numbered separately in each type’s array order. Longer text exceeds this Tool’s accepted length. Minimum length `1` Maximum length `20000`
Output length`body.duration`	`integer \| string`	Optional	Optional integer from 2 to 30 seconds, default 5, or the string auto for model-selected length. With reference video, total input-video duration plus output duration must not exceed 30 seconds. Use an explicit length to state the desired clip duration; -1 is not accepted here. Default `5`
Generated-video watermark`body.watermark`	`boolean`	Optional	Optional boolean, default false. true requests an AI Generated watermark. Default `false`
Output resolution tier`body.resolution`	`string`	Optional	Optional 480P, 720P (default) or 1080P. Sets the requested output resolution tier; review the actual completed video. Default `720P` Options `480P, 720P, 1080P`
Prompt rewriting`body.prompt_extend`	`boolean`	Optional	Optional boolean, default true. Rewrites the description and adds processing time, without guaranteeing better fidelity. It must be true when a file or webpage link is supplied. Default `true`

View raw JSON

{
  "type": "object",
  "required": [],
  "properties": {
    "body": {
      "type": "object",
      "$schema": "https://json-schema.org/draft/2020-12/schema",
      "~standard": {
        "vendor": "zod",
        "version": 1,
        "jsonSchema": {}
      },
      "properties": {
        "seed": {
          "type": "integer",
          "maximum": 2147483647,
          "minimum": 0,
          "description": "Optional integer from 0 to 2147483647. Helps reuse a random starting value, without guaranteeing identical motion, subjects or videos."
        },
        "audio": {
          "type": "boolean",
          "default": true,
          "description": "Optional boolean, default true. true enables an output audio track; false requests silent video. It does not promise that the original source soundtrack will be preserved."
        },
        "media": {
          "type": "array",
          "items": {
            "type": "object",
            "required": [
              "type",
              "url"
            ],
            "properties": {
              "url": {
                "type": "string",
                "minLength": 1,
                "description": "Required nonempty public HTTPS URL; no HTTP, base64 or local file path. Images: JPG/JPEG, PNG without alpha, BMP or WEBP, ≤20 MB, sides 240–8000 pixels, aspect 1:8–8:1. Videos: MP4/MOV, ≤100 MB, 1–15 seconds, ≥16 fps, sides 240–4096 pixels, aspect 1:8–8:1. Audio: WAV/MP3, ≤15 MB, 1–15 seconds. Files: docx/xlsx/pptx/pdf/txt/md, ≤100 MB and up to 50 pages where applicable. A link must be a public webpage requiring no login."
              },
              "type": {
                "enum": [
                  "first_frame",
                  "last_frame",
                  "reference_image",
                  "reference_video",
                  "reference_audio",
                  "file",
                  "link"
                ],
                "type": "string",
                "description": "Required: first_frame, last_frame, reference_image, reference_video, reference_audio, file or link. Use one first frame alone or with one last frame, never last-only or mixed with references/files/links. References may combine: up to 10 images, 5 videos and 5 audio clips. They may also use one file or one link; file and link are mutually exclusive, with at most one of each."
              }
            },
            "additionalProperties": false
          },
          "description": "Supply an eligible nonempty array if there is no prompt; otherwise optional. Every entry needs type and url. Image, video and audio references are ordered separately by type. Each reference clip is 1–15 seconds, and this Tool limits combined video-and-audio input duration to 15 seconds. For text-only generation, omit media."
        },
        "ratio": {
          "enum": [
            "adaptive",
            "16:9",
            "9:16",
            "1:1",
            "4:3",
            "3:4"
          ],
          "type": "string",
          "default": "adaptive",
          "description": "Optional adaptive (default), 16:9, 9:16, 1:1, 4:3 or 3:4. adaptive lets the model choose a suitable ratio from the input and intent; 21:9 is not an exposed choice."
        },
        "prompt": {
          "type": "string",
          "maxLength": 20000,
          "minLength": 1,
          "description": "Nonempty text up to 20000 characters when supplied. At least prompt or eligible media is needed. Describe a new clip or the edit; refer to reference assets as Image 1, Video 1 and Audio 1, numbered separately in each type’s array order. Longer text exceeds this Tool’s accepted length."
        },
        "duration": {
          "anyOf": [
            {
              "type": "integer",
              "maximum": 30,
              "minimum": 2
            },
            {
              "type": "string",
              "const": "auto"
            }
          ],
          "default": 5,
          "description": "Optional integer from 2 to 30 seconds, default 5, or the string auto for model-selected length. With reference video, total input-video duration plus output duration must not exceed 30 seconds. Use an explicit length to state the desired clip duration; -1 is not accepted here."
        },
        "watermark": {
          "type": "boolean",
          "default": false,
          "description": "Optional boolean, default false. true requests an AI Generated watermark."
        },
        "resolution": {
          "enum": [
            "480P",
            "720P",
            "1080P"
          ],
          "type": "string",
          "default": "720P",
          "description": "Optional 480P, 720P (default) or 1080P. Sets the requested output resolution tier; review the actual completed video."
        },
        "prompt_extend": {
          "type": "boolean",
          "default": true,
          "description": "Optional boolean, default true. Rewrites the description and adds processing time, without guaranteeing better fidelity. It must be true when a file or webpage link is supplied."
        }
      },
      "additionalProperties": false
    }
  },
  "additionalProperties": false
}

Result schema

Output fields have not been documented yet.

Additional notes

For a video-to-video task, describe the new visual treatment and the parts of the original you want the model to use as references. The example asks for clay-animation style from your own five-second clip. It requests a silent result, so assess the new scene and motion rather than expecting the original soundtrack or unchanged subject details.

Text alone can describe a new clip; reference images, videos and audio can guide one, and an eligible document or public webpage can supply content. Number each reference type separately in the description. Documents and webpages require prompt rewriting. Inputs using a first frame, either alone or with a last frame, cannot be combined with other reference material. Check the input rules before supplying materials, then review the completed MP4 and save the temporary download promptly.

Alibaba Wan3.0 video generation documentation

Questions about this tool.

Inputs, results and details to know before running it.

Get connected

Number each reference type separately in its media-array order: Image 1, Image 2; Video 1, Video 2; Audio 1, Audio 2. The first reference video is Video 1 even if a reference image appears before it.

Generate Video from Text or References · Alibaba

Generate Video from Text or References · Alibaba

You receive

Start with your task

What to know

Inputs & outputs

Input schema

Result schema

Share a link that opens the video

Choose the look for your new clip

Watch and save your new video

Questions about this tool.

You receive

Start with your task

What to know

Inputs & outputs

Input schema

Result schema

Share a link that opens the video

Choose the look for your new clip

Watch and save your new video

Questions about this tool.

How do I refer to several reference assets in the description?

Can first and last frames be combined with other references?

Does disabling audio keep the original video soundtrack?