Video Generation
Wan Series
Alibaba Tongyi Wanxiang 2.7 video-edit edition — style transfer / content replacement / video rewriting
POST
Wan Series
Alibaba Tongyi Wanxiang 2.7 video-edit edition.
Purpose-built for video-rewriting workflows — feed in a source video plus a text instruction (and up to 4 optional style reference images), get back a new video with style transfer / content replacement / full rewrite. Unlike T2V/I2V, VideoEdit requires
After submission, poll status with
video_urls and operates on “reworking an existing video”.
Billed by resolution, with linear scaling by source-video duration.
Other tiers in the Wan 2.7 family (T2V / I2V / R2V) are not yet available; please contact sales if you need them.
Pricing
VideoEdit is billed by output-video duration; when
duration=0 follows the source video, billing is based on the source-video duration.
Examples
GET /v1/tasks/{task_id}; see Task System for details.
Available Models
Common Parameters
string
required
Fixed to
wan2.7-videoeditstring
default:"edit"
Operation type. Allowed values:
edit— video edit (default)video_edit— equivalent toeditstyle_transfer— style transfer
string
required
Edit instruction text describing the target effect (replacement / transfer / rewrite), up to 5000 characters
string[]
required
Source video URL array (single element), required:
- Base64 not currently supported
- Public download URL required
string[]
Style / content reference image URL array, up to 4
string
default:"adaptive"
Frame aspect ratio. Allowed values:
16:9— landscape widescreen9:16— portrait1:1— square4:3— landscape3:4— portraitadaptive— follow source-video ratio (recommended)
string
default:"720p"
Output resolution. Allowed values:
720p1080p
integer
default:"0"
Video duration in seconds, range
0-10:0— follow source-video duration (recommended)1-10— specify output duration
string
Webhook callback URL, invoked when the task reaches a terminal state. See Callback Mechanism
string[]
Subscribed callback event types; defaults to terminal events
Model-specific Parameters
boolean
default:"false"
Add an AI-generated watermark to the bottom-right corner
boolean
default:"true"
Smart prompt rewriting — the backend auto-expands the edit instruction toward the target style
object
Additional parameter object. Commonly used
audio_setting controls the output audio-track strategy:auto— auto-generate audio matching the edited video (default)origin— keep the source video’s original audio track
Resource Limits
Related Docs
- Task System Reference — task state machine / polling cadence / async push
- Request & Response Format — common error codes / headers / rate limits
- Authentication — API key signup and usage
Wan Series