# Video Utils - Core.Today AI API

> FFmpeg-powered video utilities (20M+ runs) — convert to mp4/gif, extract audio as mp3, or dump zipped frames, all with one task parameter. The Swiss-army knife for media pipelines.

- **Provider**: FFmpeg
- **Model ID**: nicolascoutureau/video-utils
- **Category**: Video Generation
- **Credits**: 5 per operation
- **Speed**: Fast
- **Quality**: Standard

## Features
- Convert any video to mp4 or gif
- Extract audio as mp3
- Dump frames as a zip (with fps control)
- 7 credits per operation

## Use Cases
- Format normalization in pipelines
- GIF previews for social/UI
- Audio extraction before Whisper transcription
- Frame sampling for vision models

## API Endpoint

Base URL: https://api.core.today/v1
Create Prediction: POST /predictions
Get Status: GET /predictions/{job_id}

## Authentication
Header: X-API-Key: YOUR_API_KEY

## Input Parameters

### Required
- **task**: string - Operation to perform
  Options: convert_input_to_mp4, convert_input_to_gif, extract_video_audio_as_mp3, zipped_frames_to_mp4, zipped_frames_to_gif, extract_frames_from_input, reverse_video, bounce_video, create_image_preview, create_preview_video, get_video_metadata, reencode_audio_for_fast_seeking, encode_for_insta, reencode_for_web
- **input_file**: string - Input video/media file URL

### Optional
- **fps**: integer (default: 0) - Frame rate for gif/frames tasks (0 = source fps)

## Examples

### Extract Audio for Transcription
Pull the mp3 track to feed into Whisper
```json
{
  "model": "nicolascoutureau/video-utils",
  "input": {
    "input_file": "https://example.com/meeting-recording.mp4",
    "task": "extract_video_audio_as_mp3"
  }
}
```

## Response Format
```json
{
  "job_id": "abc123",
  "status": "pending | processing | completed | failed",
  "result": "URL or data (when completed)"
}
```

## Usage Flow
1. POST /predictions with model and input -> receive job_id
2. GET /predictions/{job_id} -> poll until status is completed or failed
3. Result contains output URL(s)

## Tips
- Chain extract_video_audio_as_mp3 → openai/whisper for video transcription
- zipped_frames with fps 1 gives one frame per second for vision analysis
- GIF conversion at low fps (8-12) keeps file sizes social-friendly

## Documentation
https://replicate.com/nicolascoutureau/video-utils