Generate Loop Video

Generate a looping video from an uploaded image. The complete workflow includes:

1. Upload an image file
2. Generate video from the image
3. Detect blink points in the video
4. Create a seamlessly looping video

This endpoint initiates the generation process and returns a task ID as a string. Use this task ID to poll the status endpoint.

Supported image formats: jpg, jpeg, png, gif, webp, bmp

Response: Returns a task ID string (e.g., “853377020565258316”) that you should use to query the generation status.

POST /api/open/v1/generate/generateLoopVideo

Generate Loop Video

cURL

curl --request POST \
  --url https://api.navtalk.ai/api/open/v1/generate/generateLoopVideo \
  --header 'Content-Type: multipart/form-data' \
  --header 'license: <api-key>' \
  --form image='@example-file' \
  --form 'prompt=Natural head movement, subtle smile' \
  --form 'negative_prompt=Blurry, distorted, unnatural' \
  --form beforeSeconds=0.5 \
  --form afterSeconds=0.5 \
  --form maxPollingTime=300

200

Example

{
  "code": 200,
  "message": "SUCCESS",
  "data": "853377020565258316"
}

Authorizations

license · string · header · required

NavTalk API license key (e.g. sk_navtalk_...)

Body

multipart/form-data

image · file · required

The image file to generate video from

prompt · string

Optional generation prompt to guide the video style

Example:

"Natural head movement, subtle smile"

negative_prompt · string

Optional negative prompt to avoid unwanted features

Example:

"Blurry, distorted, unnatural"

beforeSeconds · number<float>

Seconds to keep before the detected blink point (default: 0.5)

Example:

0.5

afterSeconds · number<float>

Seconds to keep after the detected blink point (default: 0.5)

Example:

0.5

maxPollingTime · integer

Maximum polling time in seconds (default: 300)

Example:

300

Response

200

application/json

Video generation task created successfully

code · integer

Example:

200

message · string

Example:

"SUCCESS"

data · string

Task ID for polling generation status. Use this ID to query the generation status via the status endpoint.

Example:

"853377020565258316"