Detailed Parameters

Version History

Date	Update Summary
2026-10-16	Added the offline video editing endpoint description: create a task via `POST /ent/s_editing/offline`, query the task and output, and error codes.

Create a Task

Submit a source video together with a set of scene switches arranged along the video timeline, and asynchronously generate an MP4. The output resolution is 540p, 24fps; width/height and orientation match the source video; the output duration equals the source video duration (it is neither padded nor truncated).

Request example

curl -X POST "$BASE_URL/ent/s_editing/offline" \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vidu-s2-editing",
    "video": "https://example.com/input/source.mp4",
    "actions": [
      {"time_s": 0,   "image_uri": "https://example.com/input/style.jpg",  "editing_type": "style_transfer"},
      {"time_s": 1.5, "image_uri": "https://example.com/input/clothes.png","editing_type": "virtual_tryon"}
    ]
  }'

Response example

{
  "task_id": 900000000000000101,
  "state": "created",
  "model": "vidu-s2-editing",
  "credits": 0,
  "created_at": "2026-10-16T10:00:00Z",
  "resolution": "540p"
}

Request Parameters

modelstring

Model identifier, defaults to vidu-s2-editing

videostringRequired

Required. Source video address (an HTTP(S) GET URL or an ssuploadid); base64 is not supported. Must be MP4 / H.264, 540p (short side 540 px), 24fps, ≤1 GiB, ≤30 minutes.

actionsobject[]Required

Required. List of scene switches, 1–64 items; see “actions[] Object Fields” below.

watermarkboolOptional

Optional. Whether to add a watermark, defaults to true.

wm_positionenum

Position of the “AI-generated content” label in the frame; only effective when watermark=true. Options: top_left, top_right, bottom_left, bottom_right; default: bottom_right. Note: this field controls the position of the AI-generated compliance label; the Vidu brand watermark is fixed in the bottom-right corner and is not affected by this field. An invalid enum value causes the request to return 400; when omitted, the default position is used.

wm_urlstringOptional

Optional. Custom watermark image URL (HTTP(S)).

callback_urlstringOptional

Optional. Callback URL for the result.

payloadstringOptional

Optional. Pass-through field, returned as-is in the callback.

meta_datastringOptional

Optional. Pass-through metadata.

actions[] Object Fields

time_snumber

The time, in seconds (decimals allowed), on the source video timeline at which the switch is triggered. The first item must be 0, and subsequent items must strictly increase.

image_uristring

The reference image address for this scene (an HTTP(S) URL, an ssuploadid, or a base64 data URI).

editing_typeenumOptional

Optional; the scene type. See “editing_type Enum Values” below; defaults to style_transfer when omitted.

editing_type Enum Values

Value	Scenario	Reference Image Meaning
`style_transfer`	Style transfer	Target visual style
`virtual_tryon`	Virtual try-on	Target garment
`subject_replacement`	Subject replacement	Target subject
`background_replacement`	Background replacement	Target background

Rules:

actions must have 1–64 items; a non-zero first time_s, a non-strictly-increasing sequence, or an empty image_uri all return 400. When editing_type is omitted or defaults to unspecified, it is treated as style_transfer; an unknown value returns 400.
time_s marks the moment the switch is triggered; the moment it actually takes effect in the output may be slightly later (roughly 1–2 seconds), and frame-exact timing is not guaranteed.
An item whose trigger time is ≥ the output duration is ignored (it does not affect task success); a new scene replaces the current one and persists until the next item or the end of the output.

Response Parameters

Field	Description
`task_id`	Task ID.
`state`	State; `created` on successful creation.
`model`	Model echoed back.
`credits`	Returns `0` at creation time; the actual deduction happens within the task workflow, based on source video duration × 1.5, and is reflected when querying the task.
`payload`	Pass-through value echoed back.
`created_at`	Creation time.
`resolution`	Output resolution (fixed at 540p).

Query the Task and Retrieve the Output

Task status: GET /ent/v2/tasks (can be filtered by task ID); terminal states are success / failed.
Output: GET /ent/v2/tasks/{task_id}/creations, returns the output video address (plus cover/last-frame images).

Error Codes

Scenario	HTTP	Description
`actions` structure invalid	400	Returns the specific field error.
Missing `video`	400	`video` is required.
`model` invalid	400	The model is not registered.
Insufficient credits	402	`INSUFFICIENT_CREDIT`
Source video does not meet spec / asset not downloadable / inference or upload failed	200 on creation, `failed` on polling	`err_code` is `OFFLINE_EDITING_*`
Input asset or output failed moderation	200 on creation, `failed` on polling / output moderation failure	—

Pricing

Billed by video duration: 1 credit per second (RMB 0.03125), with the duration rounded up from the actual duration of the input video.
A task cannot be initiated when the balance falls below 45 credits (about RMB 1.4).
When credits are deducted: the create endpoint only performs a balance pre-check and does not deduct credits; once the task enters the queue, credits are pre-deducted in one go based on the estimated duration. If the task fails or produces no output, the unconsumed portion is refunded automatically; the final deduction is the credits value returned by the query endpoint (0 when failed).