# Pruna P-Video Animate - Core.Today AI API

> Pruna AI's motion-transfer video model — animates a reference image with the motion and audio of a source video. Optimized for speed and cost (about 5.24s of generation per 1s of output video).

- **Provider**: Pruna AI
- **Model ID**: prunaai/p-video-animate
- **Category**: Video Generation
- **Credits**: 1050 per 15s 720p video — billed per second of the source video (720p 70/s, 1080p 140/s)
- **Speed**: Fast
- **Quality**: Standard

## Features
- Motion + audio transfer: the reference subject performs the source video's movement
- Fast and cheap — about 5.24 seconds of generation per 1 second of output
- 720p (~1 MP) or 1080p (~2 MP) output, aspect ratio preserved from the source
- Optional instruction prompt to steer how the subject is animated
- Turbo mode for faster generation at slightly lower quality

## Use Cases
- UGC ad variations — one brand asset animated with dozens of motion references
- Meme remixes dropping new subjects into well-known clips
- Scene recasting: keep the shot, swap the subject
- Game and avatar content from character art plus reference motion
- Music video b-roll from existing performance footage

## 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
- **video**: string - Source RGB video (.mp4): motion + audio source. The output follows its duration and aspect ratio — billing is per second of this video
- **image**: string - Reference image of the subject to animate

### Optional
- **instruction_prompt**: string - Further instruction on how the reference subject should be animated using the motion of the input video
- **resolution**: string (default: 720p) - Target ~megapixel budget: 720p ≈ 1 MP (70 credits/sec), 1080p ≈ 2 MP (140 credits/sec). Aspect ratio preserved
  Options: 720p, 1080p
- **target_fps**: string (default: original) - Target FPS for the working video
  Options: original, 24, 48
- **save_audio**: boolean (default: true) - Mux the source audio into the output video
- **ignore_audio**: boolean (default: false) - Don't condition generation on the source audio. If save_audio is true, the audio is still muxed into the final video
- **turbo**: boolean (default: false) - Faster generation for slightly lower quality
- **seed**: integer - Random seed. Leave blank for random

## Examples

### Scene Recasting
Swap the subject of a clip while keeping motion and audio
```json
{
  "model": "prunaai/p-video-animate",
  "input": {
    "video": "https://example.com/driver.mp4",
    "image": "https://example.com/subject.jpg",
    "resolution": "720p",
    "save_audio": true
  }
}
```

### Directed Animation
Add an instruction for how the subject should move
```json
{
  "model": "prunaai/p-video-animate",
  "input": {
    "video": "https://example.com/action.mp4",
    "image": "https://example.com/character.jpg",
    "instruction_prompt": "The man throws the swords to the floor, then performs the split",
    "resolution": "720p"
  }
}
```

## 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
- Billing follows the source video's length — trim the source to just the motion you need before generating
- 720p is half the price of 1080p and works well for most cases; use 1080p only when you need extra detail
- Strong, well-lit motion in the source transfers better than subtle or occluded movement
- Use a front-facing reference image — heavy angles or occlusion hurt identity preservation
- The output uses the source video's aspect ratio, so frame the source the way you want the result framed

## Documentation
https://replicate.com/prunaai/p-video-animate