​

Video Models Overview

​

Available Models

​

Kling AI (Kuaishou)

• Kling AI Pro: Professional-grade video generation with 1080p output
• Kling AI Standard: Balanced quality and speed for most use cases

​

Runway ML

• Gen-3 Alpha: Latest generation with improved motion and consistency
• Gen-2: Reliable video generation with good quality

​

Luma AI

• Dream Machine: High-quality video generation from text and images
• Dream Machine Turbo: Faster generation with reduced quality

​

Pika Labs

• Pika 1.5: Creative video generation with artistic styles
• Pika Base: Standard video generation model

​

Model Capabilities

​

Video Generation API

POST /v1/videos/generations

​

Text-to-Video Example

curl -X POST "https://api.anyapi.ai/v1/videos/generations" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kling-ai-pro",
    "prompt": "A serene mountain landscape at sunset with flowing clouds",
    "duration": 5,
    "resolution": "1080p",
    "fps": 24
  }'

​

Image-to-Video Example

import requests

# Upload image file
files = {'image': open('source_image.jpg', 'rb')}

data = {
    'model': 'runway-gen3',
    'prompt': 'The person in the image starts walking forward',
    'duration': 3,
    'motion_strength': 0.7
}

response = requests.post(
    "https://api.anyapi.ai/v1/videos/generations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    files=files,
    data=data
)

​

Response Format

{
  "id": "vid_abc123",
  "object": "video",
  "created": 1589478378,
  "model": "kling-ai-pro",
  "status": "processing",
  "url": null,
  "thumbnail_url": null,
  "duration": 5,
  "resolution": "1080p",
  "fps": 24
}

​

Video Status API

GET /v1/videos/{video_id}

import requests
import time

def wait_for_video(video_id):
    while True:
        response = requests.get(
            f"https://api.anyapi.ai/v1/videos/{video_id}",
            headers={"Authorization": "Bearer YOUR_API_KEY"}
        )
        
        video = response.json()
        
        if video['status'] == 'completed':
            return video['url']
        elif video['status'] == 'failed':
            raise Exception(f"Video generation failed: {video.get('error')}")
        
        time.sleep(10)  # Wait 10 seconds before checking again

​

Model Comparison

​

Advanced Features

​

Motion Control

{
  "model": "kling-ai-pro",
  "prompt": "A calm lake with gentle ripples",
  "motion_strength": 0.3,
  "duration": 5
}

​

Aspect Ratios

• 16:9: 1920x1080, 1280x720 (landscape)
• 9:16: 1080x1920, 720x1280 (portrait/mobile)
• 1:1: 1024x1024 (square/social media)
• 4:3: 1024x768 (traditional)

​

Camera Movement

{
  "prompt": "A beautiful garden with camera slowly zooming in on a flower",
  "camera_movement": "zoom_in",
  "motion_strength": 0.5
}

​

Prompt Engineering

​

Best Practices

1. Be specific about motion: “slowly”, “quickly”, “gently flowing”
2. Describe camera movement: “camera pans left”, “zooming out”
3. Set the scene context: lighting, time of day, weather
4. Specify style: “cinematic”, “documentary style”, “artistic”

​

Example Prompts

​

Pricing

• Image-to-video conversions
• Extended duration (>5 seconds)
• Higher frame rates (>24fps)

​

Rate Limits

​

Technical Specifications

​

Supported Input Formats

• Images: PNG, JPEG, WebP (up to 20MB)
• Video: MP4, MOV, AVI (for video extension)

​

Output Formats

• Video: MP4 (H.264 codec)
• Audio: AAC (if applicable)
• Quality: Variable bitrate optimization

​

Processing Times

• Text-to-video: 2-10 minutes depending on duration
• Image-to-video: 1-5 minutes
• Queue times: May vary during peak usage

​

Content Policy

• No harmful, violent, or inappropriate content
• Respect copyright and intellectual property
• No generation of real people without consent
• Commercial use allowed with proper licensing

​

Common Use Cases

​

Error Handling

​

Getting Started