AI Video Generation

One unified /api/v2/video/generate endpoint; switch the model via the model field (tested live)

Video generation is one unified endpoint POST /api/v2/video/generate; the model field selects the underlying model (Kling / Sora / Veo / Seedance / Vidu / Wan / SkyReels / Hailuo / Grok). Text-to-video or image-to-video. Async task; poll for status after submitting.

Base URL: https://api.aiclonevoicefree.com (same as audio) Auth: Authorization: Bearer sk_... Tested live: Kling and Seedance both produce real videos.

`POST /api/v2/video/generate`

Field	Type	Required	Notes
`prompt`	string	✅	Text prompt
`model`	string	⬜	Underlying model, default `kling-v3` (table below)
`duration`	int	⬜	Seconds, default 5
`mode`	string	⬜	Kling: `std` / `pro` / `4k`
`resolution`	string	⬜	Sora/Seedance/Vidu/Wan/SkyReels: `480p` / `720p` / `1080p` …
`aspect_ratio`	string	⬜	e.g. `16:9` / `9:16` / `1:1`
`image_urls`	string[]	⬜	Image-to-video (first/last frame)
`first_frame_url` / `last_frame_url`	string	⬜	First/last frame (alternative to image_urls)
`negative_prompt`	string	⬜	Negative prompt (some models)
`audio`	bool	⬜	Generate audio (some models)
`video_urls`	string[]	⬜	Reference videos (Seedance/SkyReels)
`seed`	int	⬜	Random seed

Model matrix & billing (video credits)

`model`	Billing
`kling-video-o1` / `kling-v3` / `kling-v3-omni`	Per second: std 33 / pro 44 / 4k 210 (v3 higher with audio)
`sora-2` / `sora-2-pro`	Per second: sora-2 720p=34; pro 720p=102 / 1024p=171 / 1080p=239
`veo3.1-fast` / `veo3.1-quality`	Flat 75 / 170 (+20 image-to-video)
`doubao-seedance-1-5-pro` / `2.0` / `2.0-fast`	Per second by resolution / video-ref, 10–168
`viduq3` / `viduq3-pro` / `viduq3-turbo` / `viduq3-mix`	Per second by resolution, 15–60
`wan-2.7`	Per second, 720P=32 / 1080P=52
`skyreels-v4-std` / `skyreels-v4-fast`	Per second by resolution / ref-video, 30–235
`MiniMax-Hailuo-02`	Flat, 6s=400 / 10s=600
`grok-video-3` / `grok-video-3-10s`	Flat 18 / 36 (+20 image-to-video)

Cost = ceil(duration × per-second rate) (except flat models). 402 at submit if video credits are insufficient; auto-refunded on failure.

Example

curl -X POST https://api.aiclonevoicefree.com/api/v2/video/generate \
  -H "Authorization: Bearer sk_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "a corgi running under cherry blossoms, cinematic",
    "model": "kling-v3",
    "mode": "std",
    "duration": 5,
    "aspect_ratio": "16:9"
  }'

Response 202

{
  "task_id": "7ee33f8d8d094c2b903c882d5eb7dfe1",
  "status": "pending",
  "capability": "video",
  "action": "generate",
  "model": "kling-v3"
}

Getting the result

Poll GET /api/v2/tasks/{task_id}; when done:

{
  "status": "completed",
  "capability": "video",
  "_type": "video.generate",
  "videoUrl": "https://.../result.mp4"
}

Model-specific parameters live on each model's page (e.g. Kling). Video generation usually takes several minutes.

AI Video Generation

POST /api/v2/video/generate

Model matrix & billing (video credits)

Example

Getting the result

On this page

`POST /api/v2/video/generate`