Kling Identify Face

Identify Face is the synchronous prerequisite for Kling Lip Sync. It checks the source video, detects faces, and returns the session_id, face_id, and time range required by the lip-sync request.

Input options

Provide exactly one source:

Field	Description
`video_url`	Public URL of an MP4 or MOV file
`video_id`	ID of a Kling-generated video created within the last 30 days

For video_url, the video must be 2–60 seconds, no more than 100 MB, 720p or 1080p, and between 512 px and 2,160 px on each side. Use a video with a clear, visible human face and minimal occlusion.

Identify a face

curl --request POST \
  --url https://www.anyfast.ai/kling/v1/videos/identify-face \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --header 'Content-Type: application/json' \
  --data '{
    "video_url": "https://example.com/person-speaking.mp4"
  }'

import requests

response = requests.post(
    "https://www.anyfast.ai/kling/v1/videos/identify-face",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "Content-Type": "application/json",
    },
    json={"video_url": "https://example.com/person-speaking.mp4"},
)
response.raise_for_status()
result = response.json()

session_id = result["data"]["session_id"]
face = result["data"]["face_data"][0]
print(session_id, face["face_id"], face["start_time"], face["end_time"])

Read the response

The response is synchronous. You do not need to poll it.

Field	Description
`data.session_id`	Session that binds the source video to its face analysis
`data.face_data[]`	Detected faces and their usable time intervals
`data.face_data[].face_id`	Face ID to pass as `face_choose[].face_id`
`data.face_data[].face_image`	Preview image of the detected face
`data.face_data[].start_time`	Start of the usable face interval, in milliseconds
`data.face_data[].end_time`	End of the usable face interval, in milliseconds
`data.final_unit_deduction`	Final Kling unit deduction for face identification

When several faces are returned, select the one whose time interval overlaps the speech segment you plan to insert. Kling Lip Sync currently accepts one selected face.

Next: Kling Lip Sync

Use the session ID and face ID to create a lip-sync task.

Identify Face API Reference

Review the complete request and response schema.

​Input options

​Identify a face

​Read the response

Next: Kling Lip Sync

Identify Face API Reference

Input options

Identify a face

Read the response