> ## Documentation Index
> Fetch the complete documentation index at: https://docs.kymaapi.com/llms.txt
> Use this file to discover all available pages before exploring further.

# MiniMax Voice Clone

> Clone a voice from a 10s-5min reference recording. Returns a voice_id usable in /v1/audio/speech with any MiniMax HD or Turbo SKU.

## Overview

`minimax-voice-clone` is a voice-cloning service, not a direct TTS model. Upload reference audio of someone speaking (10 seconds to 5 minutes) and get back a `voice_id`. That ID then becomes a usable voice on [`/v1/audio/speech`](/api-reference/audio-speech) with any MiniMax voice model.

Right for brand voice cloning when you have voice talent, character voices for games and animation, or custom narrator profiles for podcasts.

## Specs

| Field           | Value                                                   |
| --------------- | ------------------------------------------------------- |
| Model ID        | `minimax-voice-clone`                                   |
| Creator         | MiniMax                                                 |
| Best for        | Brand voice cloning, character voices, custom narrators |
| Reference audio | 10 sec – 5 min, MP3/WAV/M4A, max 10 MB                  |
| Output          | `voice_id` (kyma-namespaced)                            |
| Pricing mode    | Per call (flat)                                         |

## Pricing

|                  |          Cost          |
| ---------------- | :--------------------: |
| Per cloned voice | \$2.10 flat (one-time) |

The `voice_id` is then free to use in `/v1/audio/speech` — only the speech calls themselves bill at MiniMax's per-character rate.

## Use this when

* You have voice talent and want their voice usable across all your TTS workflows.
* You're building a character cast for a game or animation.
* You want a single brand voice spanning multiple languages and use cases.

## Pick something else when

* You don't have a reference recording → [`minimax-voice-design`](/models/minimax-voice-design) (\$4.20/voice from text description).
* You're fine with stock MiniMax voices → skip cloning, browse system voices via [`GET /v1/audio/voices`](/api-reference/audio-voices) and use directly.

## Example

```bash theme={null}
# Step 1: clone the voice
curl -X POST https://kymaapi.com/v1/audio/voice-clone \
  -H "Authorization: Bearer $KYMA_API_KEY" \
  -F file=@reference.mp3 \
  -F name="brand-narrator"
# → { "voice_id": "kyma_3f8e2a1b4c9d7e60", ... }

# Step 2: use the voice
curl -X POST https://kymaapi.com/v1/audio/speech \
  -H "Authorization: Bearer $KYMA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-speech-hd",
    "input": "Welcome to the show.",
    "voice_id": "kyma_3f8e2a1b4c9d7e60"
  }' \
  --output intro.mp3
```

## Ownership

Cloned voice IDs are gated per Kyma user. If user A passes user B's `voice_id` to `/v1/audio/speech`, the request returns `403 voice_not_owned`.

## See also

* [`POST /v1/audio/voice-clone`](/api-reference/audio-voice-clone) — endpoint reference
* [`POST /v1/audio/voice-design`](/api-reference/audio-voice-design) — generate from text instead