POST
Quote a video clone

Authorizations

Authorization
string
header
required

Provide your API key in the Authorization header using Bearer auth

Body

application/json

Creates a video clone task and starts revision 1 (POST /v1/videos/clone). Give exactly one of source_file_id and source_url.

prompt
string
required

What to change, e.g. replace the main character or the setting

Maximum string length: 4000
Example:

"把主角换成参考图里的橘猫,背景改成海边"

understand_model
string
required

Text model that watches the reference video and writes the plan; it must accept video input

Minimum string length: 1
Example:

"google/gemini-3-pro-preview"

image_model
string
required

Image model for subject sheets and frames

Minimum string length: 1
Example:

"openai/gpt-image-2"

video_model
string
required

Video model for the shot clips

Minimum string length: 1
Example:

"bytedance/seedance-1.5-pro"

source_file_id
string

ID of the reference video uploaded to this project (mp4, mov or webm, up to 5 minutes)

Example:

"file-0192f1c3a7b87c3e9a4f5b6c7d8e9f01"

source_url
string

Public http(s) link to the reference video file, downloaded in the import step (up to 200 MB and 5 minutes; web pages are not accepted)

Maximum string length: 2048
Example:

"https://example.com/reference.mp4"

reference_file_ids
string[]

IDs of reference images (up to 12), numbered in this order in the understanding and plan

Maximum array length: 12

IDs of reference images (up to 12), numbered in this order in the understanding and plan

aspect_ratio
string

auto (default, nearest to the reference video), 9:16, 16:9 or 1:1

Example:

"auto"

resolution
string

720p (default) or 1080p

Example:

"720p"

max_duration_seconds
integer<int32>

Maximum length of the final video in seconds (1–300, default 300); longer reference videos are cut down in the plan

Required range: 1 <= x <= 300
Example:

60

audio_policy
string

keep (default, original audio), mute, dub (new voiceover with voice_model) or generated (audio generated by the video model)

Example:

"keep"

voice_model
string

Speech model for audio_policy = dub (default qwen/qwen3-tts-flash)

Example:

"qwen/qwen3-tts-flash"

voice
string

Voice for the voiceover; the model default when omitted

Maximum string length: 64
Example:

"Cherry"

auto_confirm
boolean

Do not stop for confirmation after each step

budget_usd
number

Budget limit for this revision (USD)

Required range: x >= 0
Example:

20

end_user
string

End-user identifier

Maximum string length: 128
Example:

"user_42"

metadata
object

Caller metadata (up to 16 keys)

Response

OK

A quote before creating a video clone (POST /v1/videos/clone/quote).

object
string

Always video.clone.quote

Example:

"video.clone.quote"

estimated_cost_usd
number

Amount held when the task is created (model cost + workflow fee, USD)

Example:

0.12

model_cost_usd
number

Model cost part (USD)

Example:

0.06

fee_usd
number

Workflow fee part (USD)

Example:

0.06

steps
string[]

Steps in the first segment

Steps in the first segment

Example:
total_cost_usd
number

Estimate for the whole of revision 1 (USD)

Example:

4.2

source_duration_ms
integer<int64>

Reference video length the estimate is based on (ms)

Example:

15000

within_budget
boolean

When budget_usd is given: whether the first segment fits in it

Example:

true