Generate Clips

Create engaging video clips from YouTube content using our advanced multimodal AI model.

The Generate Clips endpoint allows you to automatically extract the most engaging segments from any YouTube video or public video URL. Our AI analyzes the video content across multiple dimensions including visual engagement, audio quality, semantic relevance, and viewer retention patterns to identify the most compelling clips.

POSThttps://api.joinoverlap.com/generate-clips

Overview

This endpoint initiates an asynchronous process to analyze and generate clips from your video. The process involves:

Video validation and metadata extraction
Content analysis using our multimodal AI model
Clip boundary detection and optimization
Final clip generation and processing

Authentication

All requests must include your API key in the Authorization header.

bash

Authorization: Bearer YOUR_API_KEY

Request Parameters

json

{
  "url": "string",       // YouTube or public video URL
  "minimumTime": number, // Minimum clip duration in seconds
  "maximumTime": number  // Maximum clip duration in seconds
}

url

The URL of the video to process. Must be either:

A valid YouTube URL (e.g., https://www.youtube.com/watch?v=...)

A public video URL accessible without authentication

minimumTime

The minimum duration for generated clips in seconds.

Must be between 30 and 600 seconds

Default: 60 seconds

Cannot be greater than maximumTime

maximumTime

The maximum duration for generated clips in seconds.

Must be between minimumTime and 900 seconds

Default: 300 seconds

Recommended: 2-3x minimumTime for optimal results

Response Format

json

{
  "taskId": "string",           // Unique identifier for the generation task
  "remainingHours": {
    "client": number,           // Remaining client hours
    "internal": number          // Remaining internal hours
  },
  "estimatedProcessingTime": number  // Estimated processing time in seconds
}

taskId

A unique identifier for your clip generation task. Use this ID with the Task Status endpoint to track progress.

remainingHours

Available processing hours for your account:

client: Available hours after immediate deduction for the request

internal: Hours that will remain after successful processing

estimatedProcessingTime

Estimated time in seconds to complete the clip generation. This varies based on video length and complexity.

Error Responses

json

{
  "error": "string",    // Detailed error message
  "status": number,    // HTTP status code
  "code": "string"     // Error code for programmatic handling
}

Status	Code	Description
400	INVALID_URL	The provided video URL is invalid or inaccessible
401	INVALID_API_KEY	The provided API key is invalid or expired
402	PAYMENT_REQUIRED	Your account requires payment to continue
403	INSUFFICIENT_HOURS	Not enough hours remaining in your quota
400	VIDEO_LENGTH_ERROR	Unable to determine video length or video is too short
400	VIDEO_TOO_SHORT	Video length is shorter than minimum clip length
500	SERVER_ERROR	Internal server error occurred

Video Requirements

Video must be publicly accessible without authentication
Video length must be greater than the specified minimum clip length
Maximum supported video length is 4 hours
Supported formats: MP4, MOV, AVI, WebM
Maximum file size: 2GB

Additional Endpoints

The following additional endpoints are available for managing clip generation tasks:

GET /get-successful-generations
Retrieve a list of your recent successful clip generation tasks
GET /get-clips-from-task/{taskId}
Get all clips generated from a specific task

Code Examples

bash

curl -X POST https://api.joinoverlap.com/generate-clips \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "url": "https://www.youtube.com/watch?v=example",
    "minimumTime": 60,
    "maximumTime": 300
  }'

Usage Notes & Best Practices

Processing Time

Processing takes approximately 10-15 seconds per minute of content
Videos longer than 60 minutes may take additional time for initial processing
Use webhooks for real-time notifications when processing completes

Usage & Quotas

Monthly quota of 150 processing hours
Hours are tracked in two ways:
- Client hours: Deducted immediately upon request
- Internal hours: Deducted upon successful processing
Failed processing attempts do not count against internal hours

Optimization Tips

For optimal results, provide videos between 5-60 minutes in length
Set maximumTime to 2-3x minimumTime for better clip selection
The API generates 5-10 clips per video based on content quality
Higher quality source videos result in better clip detection

Error Handling

Implement exponential backoff for rate limit errors
Store taskId values securely for status checking
Monitor remainingHours to avoid quota exhaustion
Use webhook notifications for reliable status updates