Digital Human Video

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.

Based on a single portrait image + a vocal audio clip, generate a talking, singing, or performing video with lip movements, facial expressions, and body motions synchronized to the audio. Supports real people (portrait / half-body / full-body) as well as cartoon characters.

This document covers two endpoints:

Endpoint	Model	Invocation
Digital Human Video generation	`wan2.2-s2v`	Asynchronous (create task + poll for results)
Face Detection	`wan2.2-s2v-detect`	Synchronous (returns result immediately)

Recommended workflow: Call the Face Detection endpoint first to confirm the image is compliant (check_pass: true), then submit the Digital Human Video generation task. This avoids task failures caused by non-compliant images.

1. Digital Human Video Generation (wan2.2-s2v)

Create Task

POST https://platform.dataeyes.ai/ali/api/v1/services/aigc/image2video/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 added automatically by the platform — you do not need to include it; including it is also compatible.

Request Body

Parameter	Type	Required	Default	Description
`model`	string	Yes		Fixed to `wan2.2-s2v`
`input.image_url`	string	Yes		Portrait image URL; must be a publicly accessible HTTP/HTTPS link. See "Image and Audio Requirements" below
`input.audio_url`	string	Yes		Vocal audio URL; must be a publicly accessible HTTP/HTTPS link. See "Image and Audio Requirements" below
`parameters.resolution`	string	No	480P	Output resolution tier, options `480P`, `720P`. Output keeps the same aspect ratio as the input image, with total pixels adjusted to around the selected tier (480P ≈ 310K pixels, 720P ≈ 920K pixels; e.g. a 4:5 input image at 480P may output 480×600)
`parameters.style`	string	No		Lip-sync scenario, e.g. `speech` (talking). The official documentation describes three scenarios — talking, singing, and performing — but the enum values for singing/performing are not publicly documented; verify through testing

Video Duration Rules

Generated video duration = input audio duration; it is not specified via a parameter:

The audio duration must be less than 20 seconds, otherwise the task will fail with an error;

For longer videos, split the audio into segments of less than 20 seconds each, generate them separately, then concatenate the results.

Image and Audio Requirements

Input	Requirements
Image	Format jpg/jpeg/png/bmp/webp; both width and height within [400, 7000] pixels; a clear, front-facing, single-person portrait is recommended (you can pre-check with the Face Detection endpoint)
Audio	Format wav/mp3; file smaller than 15MB, duration less than 20 seconds; must contain clear and loud vocals — removing ambient noise and background music is recommended

Request Example (480P)

curl --request POST \
  --url 'https://platform.dataeyes.ai/ali/api/v1/services/aigc/image2video/video-synthesis' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "wan2.2-s2v",
    "input": {
      "image_url": "https://example.com/portrait.jpg",
      "audio_url": "https://example.com/speech.mp3"
    },
    "parameters": {
      "resolution": "480P",
      "style": "speech"
    }
  }'

Request Example (720P)

curl --request POST \
  --url 'https://platform.dataeyes.ai/ali/api/v1/services/aigc/image2video/video-synthesis' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "wan2.2-s2v",
    "input": {
      "image_url": "https://example.com/portrait.jpg",
      "audio_url": "https://example.com/sing.wav"
    },
    "parameters": {
      "resolution": "720P",
      "style": "speech"
    }
  }'

Create Response Example

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

Usage and Billing

After the task succeeds, in the usage object returned by the query endpoint:

usage.duration: duration of the generated video in seconds, equal to the input audio duration — this is the billing basis;

usage.SR: output resolution tier (480 or 720); unit prices differ by tier, with 720P priced higher than 480P.

Digital Human Video is billed by the number of seconds of successfully generated video and the resolution tier; inputs are not billed; failed or errored tasks are not billed. Generation typically takes about 5–10 minutes; polling the query endpoint at roughly 15-second intervals is recommended.

2. Face Detection (wan2.2-s2v-detect)

Pre-checks whether the input image meets the portrait requirements for Digital Human Video generation (e.g. clarity, single person, front-facing), returning whether it passes and whether a human figure is detected.

POST https://platform.dataeyes.ai/ali/api/v1/services/aigc/image2video/face-detect

This is a synchronous endpoint: the detection result is returned immediately in the response, no task_id is produced, and no polling is needed.

Request Headers

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

Request Body

Parameter	Type	Required	Default	Description
`model`	string	Yes		Fixed to `wan2.2-s2v-detect`
`input.image_url`	string	Yes		URL of the image to check; must be a publicly accessible HTTP/HTTPS link. Format jpg/jpeg/png/bmp/webp; both width and height within [400, 7000] pixels

Request Example

curl --request POST \
  --url 'https://platform.dataeyes.ai/ali/api/v1/services/aigc/image2video/face-detect' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "wan2.2-s2v-detect",
    "input": {
      "image_url": "https://example.com/portrait.jpg"
    }
  }'

Response Example

Check passed:

{
  "output": {
    "check_pass": true,
    "humanoid": true
  },
  "usage": {
    "image_count": 1
  },
  "request_id": "c56f62df-xxxx-xxxx-xxxx-xxxxxx"
}

Check failed:

{
  "output": {
    "check_pass": false,
    "humanoid": false,
    "code": "xxx",
    "message": "xxx"
  },
  "usage": {
    "image_count": 1
  },
  "request_id": "c56f62df-xxxx-xxxx-xxxx-xxxxxx"
}

Response Fields

Field	Type	Description
`output.check_pass`	boolean	Detection result: `true` = passed, the image can be used for Digital Human Video generation; `false` = failed
`output.humanoid`	boolean	Whether a human figure was detected
`output.code` / `output.message`	string	Returned only when the check fails, giving the reason; absent when the check passes
`usage.image_count`	integer	Number of images checked in this request, fixed at 1 — this is the billing basis
`request_id`	string	Unique ID of this request

Face Detection is billed by the number of images checked: a successful request is billed regardless of whether the check passes; failed requests are not billed.

Query Task

Digital Human Video generation (wan2.2-s2v) is an asynchronous task; after creation, poll for status and results using task_id (Face Detection is synchronous and requires no querying):

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

See 09-Query-Task for details.

07-Digital-Human-Video

Digital Human Video#

1. Digital Human Video Generation (wan2.2-s2v)#

Create Task#

Request Headers#

Request Body#

Video Duration Rules#

Image and Audio Requirements#

Request Example (480P)#

Request Example (720P)#

Create Response Example#

Usage and Billing#

2. Face Detection (wan2.2-s2v-detect)#

Request Headers#

Request Body#

Request Example#

Response Example#

Response Fields#

Query Task#

Digital Human Video

1. Digital Human Video Generation (wan2.2-s2v)

Create Task

Request Headers

Request Body

Video Duration Rules

Image and Audio Requirements

Request Example (480P)

Request Example (720P)

Create Response Example

Usage and Billing

2. Face Detection (wan2.2-s2v-detect)

Request Headers

Request Body

Request Example

Response Example

Response Fields

Query Task