Video Generation AI API

한국어로 보기: Video 생성 AI API | View in English (current page)

kvidAI's Video Generation AI API creates high-quality videos from text, images, or reference media, specializing in K-pop and K-beauty content.

🎯 Service Overview

Supported Features

• Text-to-Video: Generate videos from text prompts (txt2vid)
• Image-to-Video: Animate an input image with a motion prompt (img2vid)
• Reference-to-Video: Generate consistent videos from reference images/videos/audio (ref2vid)
• Resolution: 480p / 720p / 1080p (model-dependent)
• Duration: typically 4–15 seconds (model-dependent)

Specialized Capabilities

• Camera angle manipulation prompts (may not be perfect)
• Various generation options and controls
• K-pop dance and K-beauty content optimization

📡 API Endpoints

Basic Information

Base URL:       https://api.kvid.ai
Authentication: api-key header
Content-Type:   application/json

The Video Generation API is asynchronous — first POST a generation request to get a job_id, then poll the status endpoint until the job completes, and finally fetch the result.

Method	Path	Purpose
POST	/ai/generation/text-to-video/generate-async	Submit text-to-video job
POST	/ai/generation/image-to-video/generate-async	Submit image-to-video job
POST	/ai/generation/reference-to-video/generate-async	Submit reference-to-video job
GET	/ai/generation/status?jobId={job_id}	Check job status (shared endpoint)
GET	/ai/generation/result?jobId={job_id}	Fetch completed result (shared endpoint)

Authentication & credit identification. Every request must send the api-key header. In addition, the AI-generation endpoints require exactly one of product_id / product_code / email in the request body to identify the credit pool to charge. Include one of them in every generate request.

A separate dev routing surface exists (api.hometip.net + /ai/generation-clone/...); this page documents the production paths on api.kvid.ai.

1. Create a text-to-video job

Python Example

import requests

url = "https://api.kvid.ai/ai/generation/text-to-video/generate-async"
api_key = "YOUR_API_KEY"

payload = {
    "product_id": "pdt_XXXXXXXXXXXX",   # or product_code / email — required
    "prompt": "A beautiful sunset over the ocean",
    "model": "veo3.1",                   # wan / seedance / veo3.1
    "function": "txt2vid",
    "resolution": "720p",                # 480p / 720p / 1080p (model-dependent)
    "duration": 4,
    "aspect_ratio": "16:9",
    "seed": 5834
}
headers = {
    "api-key": api_key,
    "Content-Type": "application/json",
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

Response:

{
  "success": true,
  "data": {
    "job_id": "job_1768540311147_4mcdv65c7",
    "status": "queued",
    "message": "Video generation job queued.",
    "estimated_time": "2-5min",
    "video_type": "text-to-video"
  }
}

2. Create an image-to-video job

import requests

url = "https://api.kvid.ai/ai/generation/image-to-video/generate-async"
api_key = "YOUR_API_KEY"

payload = {
    "product_id": "pdt_XXXXXXXXXXXX",   # or product_code / email — required
    "prompt": "windy, forest, autumn",
    "model": "wan",
    "function": "img2vid",
    "image_url": "https://your-host.example/scene.png",   # or image_file (base64)
    "resolution": "720p",
    "duration": 5,
    "aspect_ratio": "auto",              # auto follows the input image ratio
    "seed": 5834
}
headers = {
    "api-key": api_key,
    "Content-Type": "application/json",
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

3. Create a reference-to-video job

Generate a consistent video from one or more reference images, videos, and/or audio clips. Useful for keeping a character or style stable across the output.

import requests

url = "https://api.kvid.ai/ai/generation/reference-to-video/generate-async"
api_key = "YOUR_API_KEY"

payload = {
    "product_id": "pdt_XXXXXXXXXXXX",   # or product_code / email — required
    "prompt": "the character from the reference image walks through a neon city at night",
    "model": "bytedance/seedance-2.0/fast/reference-to-video",
    "function": "ref2vid",
    "image_urls": ["https://your-host.example/ref-character.png"],
    "video_urls": [],
    "audio_urls": [],
    "resolution": "720p",
    "duration": 5,             # 4–15 (integer) or "auto"
    "aspect_ratio": "auto",
    "generate_audio": True,
    "seed": 5834
}
headers = {
    "api-key": api_key,
    "Content-Type": "application/json",
}

response = requests.post(url, headers=headers, json=payload)
print(response.json())

Reference constraints: up to 9 images, up to 3 videos (≤ 15 s combined), up to 3 audio clips (≤ 15 s combined). Total references (image + video + audio) ≤ 12. If audio_urls is provided, at least one image or video reference is also required.

4. Check job status

import requests

api_key = "YOUR_API_KEY"
job_id = "job_1768540311147_4mcdv65c7"

url = f"https://api.kvid.ai/ai/generation/status?jobId={job_id}"
headers = {"api-key": api_key}

response = requests.get(url, headers=headers)
print(response.json())

Response (in-progress):

{
  "success": true,
  "data": {
    "job_id": "job_1768540311147_4mcdv65c7",
    "status": "processing",
    "prompt": "A beautiful sunset over the ocean",
    "result_url": null,
    "error_message": null
  }
}

status is one of: queued, processing, completed, failed, canceled.

Recommended polling interval for video jobs: 10–15 seconds (generation can take tens of seconds to a few minutes).

5. Fetch the completed result

import requests

api_key = "YOUR_API_KEY"
job_id = "job_1768540311147_4mcdv65c7"

url = f"https://api.kvid.ai/ai/generation/result?jobId={job_id}"
headers = {"api-key": api_key}

response = requests.get(url, headers=headers)
print(response.json())

Response:

{
  "success": true,
  "data": {
    "job_id": "job_1768540311147_4mcdv65c7",
    "status": "completed",
    "result_url": "https://cdn.kvid.ai/videos/job_1768540311147_4mcdv65c7.mp4",
    "created_at": "2026-05-27T09:00:00.000Z",
    "prompt": "A beautiful sunset over the ocean",
    "width": 1280,
    "height": 720,
    "size": 5242880,
    "file_size": 5242880,
    "type": "text-to-video",
    "used_credit": 54
  }
}

📋 Schema

Common request fields

Field	Type	Required	Description
product_id / product_code / email	string	✅ (one of)	Identifies the credit pool to charge
prompt	string	✅	Text prompt guiding generation
model	string	–	Model identifier (wan, seedance, veo3.1, …). Default: wan
function	string	–	txt2vid / img2vid / ref2vid (matches the endpoint)
negative_prompt	string	–	Elements to exclude
resolution	string	–	480p / 720p / 1080p. Default: 480p (model-dependent upper bound)
duration	integer | string	–	Clip length in seconds; alternative to num_frames. ref2vid accepts 4–15 or "auto"
aspect_ratio	string	–	16:9 / 9:16 / 1:1 / auto
seed	integer	–	Random seed for reproducibility

Image-to-Video specific

Field	Type	Required	Description
image_url / image_file	string	✅	Start frame — HTTPS URL (image_url) or base64 (image_file)

Reference-to-Video specific

Field	Type	Default	Description
image_urls	string[]	[]	Reference images (max 9)
video_urls	string[]	[]	Reference videos (max 3, ≤ 15 s combined)
audio_urls	string[]	[]	Reference audio (max 3, ≤ 15 s combined)
generate_audio	boolean	true	Generate audio alongside the video

V1 (self-hosted) model parameters

These apply to the self-hosted V1 model family (e.g. wan); support varies per model.

Field	Type	Default	Description
num_frames	integer	model-dependent	Frame count; alternative to duration
frames_per_second	integer	model-dependent	Output FPS
num_inference_steps	integer	model-dependent	Inference steps; higher = better quality, slower
guidance_scale	number	model-dependent	Prompt adherence strength
shift	number	model-dependent	Scheduler shift
enable_safety_checker	boolean	true	NSFW filter
enable_prompt_expansion	boolean	model-dependent	Auto-expand the prompt
acceleration	string	–	regular / high priority processing

Each model allows a different set of parameters (validated against the DB model-parameter table); unsupported fields are sanitized and ignored by the gateway backend. Model availability per resolution and exact per-model params — see Pricing.

⚠️ Errors

Code	HTTP	Meaning
MISSING_PARAMETERS / INVALID_PARAMETERS	400	Missing prompt/image or invalid parameter
INSUFFICIENT_CREDIT	402	Not enough credits
CONCURRENT_LIMIT	429	Too many concurrent jobs
—	403	api-key invalid
JOB_NOT_FOUND	404	jobId not found (or not owned by caller) — result endpoint
JOB_NOT_COMPLETED	400	Status is still queued/processing — result endpoint
JOB_FAILED	400	Status is failed; see error_message from the status endpoint

🎬 Usage Examples

1. Hiker with Backpack Video

Prompt: A video of a man hiking with a backpack. The bag must be the main subject. Walking slowly

2. Vacuum Cleaner Usage Video

Prompt: Video of cleaning with a vacuum cleaner. slow movement. low angle

3. Food Promotional Video

Prompt: Remove the cooking effect and only add camera movement. highlight the food in Zoom format. promotional video for this food

4. Korean Traditional Pavilion Video

Prompt: A traditional Korean pavilion by a lotus pond, with two small dogs (a white poodle and a brown shiba inu) joyfully running along the wooden walkway.

5. Tiger Image-to-Video Conversion

Prompt: The tiger briefly pulls back its tongue, blinks, and tilts its head slightly. [Low-angle close-up shot]

⚠️ Limitations & Notes

Technical Limitations

• Duration: short clips (4–15 s typical, model-dependent)
• Resolution: depends on selected model (see Pricing)
• Camera Angles: camera angle manipulation prompts may not always work accurately

Optimization Tips

• Specific Prompts: provide detailed, clear descriptions
• Camera Angles: use directives like [Low-angle], [Over-the-shoulder shot] when needed
• Appropriate Resolution: match resolution to your delivery channel

🔗 Related Links

• Create an API key
• Buy credits
• Pricing
• Talk-V2V (Lip-Sync) API
• Voice (TTS) API

📞 Support & Contact

For questions or assistance:

• Email: [email protected]
• Discord: kvidAI Community

---

Language: English (current page) | 한국어