# MiniMax Voice Cloning - Core.Today AI API

> Clone any voice from a 10-second to 5-minute audio sample. Returns a custom voice_id you can pass to MiniMax speech models, plus a preview clip synthesized with the cloned voice.

- **Provider**: MiniMax
- **Model ID**: minimax/voice-cloning
- **Category**: Audio & TTS
- **Credits**: 6980 per clone (one-time; returns a reusable voice_id)
- **Speed**: Medium
- **Quality**: High

## Features
- Voice cloning from as little as 10 seconds of audio
- Cloned voice_id works with MiniMax speech models
- Preview audio clip included in the result
- Optional noise reduction for noisy samples
- Optional volume normalization

## Use Cases
- Brand ambassador and narrator voice reuse
- Consistent character voices for games and animation
- Personalized voice assistants
- Audiobook narration in a specific voice
- Localized dubbing with a consistent voice identity

## 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
- **voice_file**: string - Voice file to clone. MP3, M4A, or WAV, 10s-5min, under 20MB

### Optional
- **model**: string (default: speech-02-turbo) - The text-to-speech model to train the cloned voice for
  Options: speech-2.6-turbo, speech-2.6-hd, speech-02-turbo, speech-02-hd
- **accuracy**: number (default: 0.7) - Text validation accuracy threshold (0-1)
- **need_noise_reduction**: boolean (default: false) - Enable noise reduction if the voice file has background noise
- **need_volume_normalization**: boolean (default: false) - Enable volume normalization

## Examples

### Clone a Narrator Voice
Clone a clean studio recording and get a reusable voice_id
```json
{
  "model": "minimax/voice-cloning",
  "input": {
    "voice_file": "https://example.com/narrator-sample.wav",
    "model": "speech-02-turbo",
    "accuracy": 0.7,
    "need_noise_reduction": false,
    "need_volume_normalization": false
  }
}
```

## 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
- Use a clean 30s-1min sample with a single speaker for the best clone quality
- Pass the returned voice_id as the voice_id parameter of MiniMax speech models
- Enable need_noise_reduction only when the sample has audible background noise
- Pick the model you will actually synthesize with — the clone is trained per model
- You must have permission to clone the voice — do not clone voices without consent

## Documentation
https://replicate.com/minimax/voice-cloning