Video Generation

📝 Introduction

Video generation is asynchronous: submit a task, receive a task ID, then poll for the result.

Family	Models	Create Endpoint
Google Veo	veo-3.1-generate-preview, veo-3.1-fast-generate-preview, veo-3.0-generate-001, veo-3.0-fast-generate-001	POST /v1/video/generations
Vidu	viduq3-pro, viduq3-turbo, viduq3-fast	POST /v1/video/generations
HappyHorse	happyhorse-1.0-t2v	POST /v1/video/generations
Kling	kling-v3, kling-v2-6, kling-v2-5-turbo	POST /kling/v1/videos/text2video, POST /kling/v1/videos/image2video

Authentication

Authorization: Bearer $WS_API_KEY

Unified Format (/v1/video/generations)

Create a Video Generation Task ✅

Supports text-to-video and image-to-video.

curl -X POST "http://baseurl/v1/video/generations" \
  -H "Authorization: Bearer $WS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "veo-3.1-generate-preview",
    "prompt": "A drone shot flying over a coastal city at golden hour, cinematic",
    "metadata": {
      "resolution": "720p",
      "ratio": "16:9",
      "duration": 8,
      "generate_audio": true
    }
  }'

For image-to-video, add the images field:

{
  "model": "viduq3-pro",
  "prompt": "The camera slowly zooms in while leaves fall",
  "images": ["https://example.com/first-frame.jpg"]
}

Response — a task ID:

{"task_id": "task_id"}

Request Parameters

Parameter	Type	Required	Description
model	string	Yes	Video model ID, see table above
prompt	string	Yes	Text description of the desired video
images	array	No	Reference image(s) (URL or base64) for image-to-video
metadata.resolution	string	No	e.g. 480p, 720p, 1080p (model-dependent)
metadata.ratio	string	No	Aspect ratio, e.g. 16:9, 9:16
metadata.duration	integer	No	Video duration in seconds (model-dependent)
metadata.generate_audio	boolean	No	Whether to generate audio (models that support it)

Query Task Status ✅

curl http://baseurl/v1/video/generations/task_id \
  -H "Authorization: Bearer $WS_API_KEY"

Response Example:

{
  "code": "success",
  "message": "",
  "data": {
    "task_id": "task_id",
    "action": "generate",
    "status": "SUCCESS",
    "fail_reason": "",
    "submit_time": 1775188030,
    "start_time": 1775188043,
    "finish_time": 1775188329,
    "progress": "100%",
    "data": {
      "id": "task_id",
      "status": "succeed",
      "duration": 8,
      "resolution": "720p",
      "result": {
        "content": {
          "video_url": "video download URL"
        }
      }
    }
  }
}

Task status values:

• queued: queued
• in_progress: in progress
• completed: completed
• failed: failed

Download promptly

Generated video URLs are temporary. Download and store the video on your side as soon as the task completes.

Kling Format

Text-to-Video — Create Task ✅

curl -X POST "http://baseurl/kling/v1/videos/text2video" \
  -H "Authorization: Bearer $WS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kling-v3",
    "prompt": "A cat playing piano in the garden",
    "negative_prompt": "blurry, low quality",
    "cfg_scale": 0.7,
    "mode": "pro",
    "aspect_ratio": "16:9",
    "duration": "4",
    "metadata": {
      "sound": "on"
    }
  }'

Request parameters:

Parameter	Type	Required	Description
model	string	Yes	kling-v3, kling-v2-6, kling-v2-5-turbo
prompt	string	Yes	Positive prompt describing the desired video content
negative_prompt	string	No	Negative prompt describing content to avoid
cfg_scale	float	No	Prompt adherence, range 0–1; higher follows the prompt more closely
mode	string	No	std (standard) or pro (high quality)
aspect_ratio	string	No	16:9, 9:16, 1:1
duration	string	No	Video duration in seconds, e.g. 4, 5, 10
metadata.sound	string	No	on to generate audio, off to disable

Response Example:

{
  "code": 0,
  "message": "SUCCEED",
  "request_id": "request_id",
  "data": {
    "task_id": "task_id",
    "task_status": "submitted",
    "created_at": 1775188030000,
    "updated_at": 1775188030000
  }
}

Text-to-Video — Query Task ✅

curl "http://baseurl/kling/v1/videos/text2video/task_id" \
  -H "Authorization: Bearer $WS_API_KEY"

Response Example:

{
  "code": 0,
  "message": "SUCCEED",
  "request_id": "request_id",
  "data": {
    "task_id": "task_id",
    "task_status": "succeed",
    "task_status_msg": "",
    "created_at": 1775188030000,
    "updated_at": 1775188329000,
    "task_result": {
      "videos": [
        {
          "id": "video_id",
          "url": "video download URL",
          "duration": "4"
        }
      ]
    }
  }
}

Image-to-Video — Create Task ✅

Uses an image as the first frame (optional last frame).

curl -X POST "http://baseurl/kling/v1/videos/image2video" \
  -H "Authorization: Bearer $WS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kling-v3",
    "image": "https://example.com/first-frame.jpg",
    "image_tail": "https://example.com/last-frame.jpg",
    "prompt": "A cat playing piano in the garden",
    "cfg_scale": 0.7,
    "mode": "pro",
    "duration": "4"
  }'

Additional parameters vs. text-to-video:

Parameter	Type	Required	Description
image	string	Yes	First frame image, URL or Base64
image_tail	string	No	Last frame image, URL or Base64

Image-to-Video — Query Task ✅

curl "http://baseurl/kling/v1/videos/image2video/task_id" \
  -H "Authorization: Bearer $WS_API_KEY"

Kling task status values:

• submitted: submitted
• processing: in progress
• succeed: completed
• failed: failed