Video Editing

Doc version：v1.0.0 | Last updated：2026-07-22
This platform fully supports the official Tongyi Wanxiang (Wan) video generation API. Requests and responses are transparently proxied; parameter semantics are identical to the official API.

Edit an existing video via text instructions (wan2.7-videoedit). Two usage patterns are supported:

Capability	media composition	Description
Instruction-only editing	Only 1 `video`	Modify video content directly via instructions: add/remove/modify elements, change colors, modify environment or background, convert video style, change character behavior/lines/camera work, etc.
Instruction + reference-image editing	1 `video` + up to 4 `reference_image`	Apply elements from reference images (clothing, props, etc.) to the video, or replicate motion/camera moves/effects (video transfer)

Tasks are asynchronous: task creation returns output.task_id, then poll the query endpoint for the result. Generation typically takes 1–5 minutes.

Create Task

POST https://platform.dataeyes.ai/ali/api/v1/services/aigc/video-generation/video-synthesis

Request Headers

Parameter	Type	Required	Default	Description
`Content-Type`	string	Yes	application/json	Data exchange format
`Authorization`	string	Yes		Authentication, `Bearer {API_KEY}`

The X-DashScope-Async: enable header required by the official API is automatically added by the platform — no need to include it in your request.

Request Body

Parameter	Type	Required	Default	Description
`model`	string	Yes		Model identifier, fixed to `wan2.7-videoedit`
`input`	object	Yes		Basic input information
`input.prompt`	string	No		Editing instruction prompt describing the desired elements and visual characteristics. Chinese or English, ≤ 5000 characters, automatically truncated if longer
`input.negative_prompt`	string	No		Negative prompt describing content you do not want to appear. ≤ 500 characters, automatically truncated if longer
`input.media`	array	Yes		List of media assets; elements are `{type, url}` objects. Exactly 1 `video`; up to 4 `reference_image`
`input.media[].type`	string	Yes		`video`: the video to edit (required) `reference_image`: reference image (optional)
`input.media[].url`	string	Yes		Asset location. Videos only support public URLs (HTTP/HTTPS); images support public URLs or Base64 (`data:{MIME_type};base64,{base64_data}`, supporting JPEG/PNG/BMP/WEBP)
`parameters`	object	No		Video processing parameters
`parameters.duration`	integer	No	0	Output video duration in seconds. Default 0 means the input video's duration is inherited directly, with no truncation; only pass an integer in [2, 10] when truncation is needed — the system trims the source video from second 0 to that length
`parameters.resolution`	string	No	1080P	Output resolution tier, options `720P`, `1080P`
`parameters.ratio`	string	No	Approximate ratio of the input video	Output aspect ratio, options `16:9`, `9:16`, `1:1`, `4:3`, `3:4`; if omitted, the output uses an approximate ratio based on the input video
`parameters.audio_setting`	string	No	auto	Audio setting. `auto`: the model decides intelligently based on the prompt (when the instruction involves audio descriptions the audio may be regenerated, otherwise the original audio may be kept); `origin`: force keeping the input video's original audio, no regeneration
`parameters.prompt_extend`	boolean	No	true	Whether to enable intelligent prompt rewriting. Noticeable improvement for shorter prompts, but increases processing time
`parameters.watermark`	boolean	No	false	Whether to add an "AI-generated" watermark in the bottom-right corner of the video
`parameters.seed`	integer	No	Random	Random seed, range [0, 2147483647]. A fixed seed improves reproducibility but does not guarantee identical results

duration=0 semantics (important): parameters.duration defaults to 0, meaning the output duration inherits the input video's duration; when you want the output to match the input length, no setting is needed (or pass 0). Since the actual duration can only be determined after the task completes, rely on the usage field in the query result.

resolution × ratio Output Resolution Table

Tier	Aspect ratio	Output (width × height)
720P	16:9	1280×720
720P	9:16	720×1280
720P	1:1	960×960
720P	4:3	1104×832
720P	3:4	832×1104
1080P	16:9	1920×1080
1080P	9:16	1080×1920
1080P	1:1	1440×1440
1080P	4:3	1648×1248
1080P	3:4	1248×1648

Input Asset Requirements

Video to edit (type=video):

Item	Constraint
Count	Exactly 1 (required)
Format	MP4, MOV
Duration	2–10 seconds
Resolution	Both width and height in [240, 4096] pixels
Aspect ratio	1:8 ~ 8:1
File size	≤ 100MB
Input method	Public URL (HTTP/HTTPS); Base64 not supported

Reference image (type=reference_image):

Item	Constraint
Count	Up to 4 (optional)
Format	JPEG, JPG, PNG (alpha channel not supported), BMP, WEBP
Resolution	Both width and height in [240, 8000] pixels
Aspect ratio	1:8 ~ 8:1
File size	≤ 20MB
Input method	Public URL or Base64

Request Example

curl --request POST \
  --url 'https://platform.dataeyes.ai/ali/api/v1/services/aigc/video-generation/video-synthesis' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "wan2.7-videoedit",
    "input": {
      "prompt": "Paint the exterior walls of the building in the video a vivid red",
      "media": [
        {
          "type": "video",
          "url": "https://help-static-aliyun-doc.aliyuncs.com/file-manage-files/zh-CN/20260409/dozxak/Wan_Video_Edit_33_1.mp4"
        }
      ]
    },
    "parameters": {
      "audio_setting": "origin",
      "duration": 0
    }
  }'

Response Example

{
  "output": {
    "task_status": "PENDING",
    "task_id": "0385dc79-5ff8-4d82-bcb6-xxxxxx"
  },
  "request_id": "4909100c-7b5a-9f92-bfe5-xxxxxx"
}

Response Fields

Field	Type	Description
`output.task_id`	string	Task ID, used to query the result; valid for 24 hours
`output.task_status`	string	Task status, `PENDING` on successful creation
`request_id`	string	Unique request identifier, used for tracing and troubleshooting
`code` / `message`	string	Returned only when creation fails; error code and details

Billing Notes

After the query task succeeds, the fields in usage are the basis for billing (qualitative description; see the platform pricing page for unit prices):

usage.duration: total billed duration in seconds = input video duration + output video duration (input_video_duration + output_video_duration);

usage.SR: output video resolution tier (e.g. 720, 1080), which together with duration determines the cost;

Input reference images are not billed; failed tasks incur no charges.

Query Task

Poll the task status with the task_id returned by task creation; once successful, get the video from output.video_url (valid for 24 hours).

GET https://platform.dataeyes.ai/ali/api/v1/tasks/{task_id}

See 09-Query-Task for details.

04-Video-Editing

Video Editing#

Create Task#

Request Headers#

Request Body#

resolution × ratio Output Resolution Table#

Input Asset Requirements#

Request Example#

Response Example#

Response Fields#

Billing Notes#

Query Task#