# Voice Changer

Create song covers with any RVC v2 trained AI voice from audio files.

## API Information

- **Model Slug:** realistic-voice-cloning
- **Branded URL:** https://www.eachlabs.ai/eachlabs/eachlabs/realistic-voice-cloning
- **Provider:** each::labs
- **Category:** Voice to Voice
- **Output Type:** audio
- **Status:** active
- **Version:** 0.0.1
- **Base Cost:** Per-second pricing based on provider predict_time. Rate: $0.0002475/sec from GPU tier.
- **Estimated Processing Time:** 143 seconds
- **Last Updated:** 2026-04-06
- **Interactive Demo:** https://www.eachlabs.ai/ai-models/realistic-voice-cloning

## Pricing

- **Charge Type:** dynamic
- **Pricing Details:** Per-second pricing based on provider predict_time. Rate: $0.0002475/sec from GPU tier.

### Pricing Rules

| Condition | Pricing |
| --- | --- |
| Rule 1 | Per-second pricing based on provider predict_time. Rate: $0.0002475/sec from GPU tier. |

## Input Schema

| Parameter | Type | Required | Default | Constraints | Description |
|-----------|------|----------|---------|-------------|-------------|
| song_input | string | Yes | - | audio/mp3, audio/wav | The original song file provided as input. |
| rvc_model | string | No | Squidward | Squidward,MrKrabs,Plankton,Drake,Vader,Trump,Biden,Obama,Guitar,Voilin,CUSTOM | The specific RVC model used for the function. |
| custom_rvc_model_download_url | string | No | - | - | URL to download a custom RVC model. |
| pitch_change | string | No | no-change | no-change,male-to-female,female-to-male | Alteration in the pitch of the audio. |
| index_rate | number | No | 0.5 | 0–1 | The frequency at which indexing is performed. |
| filter_radius | integer | No | 3 | 0–7 | Range of frequencies affected by the filter. |
| rms_mix_rate | number | No | 0.25 | 0–1 | Ratio of root mean square levels for mixing. |
| pitch_detection_algorithm | string | No | rmvpe | rmvpe,mangio-crepe | The method used to detect the pitch of the vocals and instruments. |
| crepe_hop_length | integer | No | 128 | - | The step size for the pitch detection process using the CREPE algorithm. |
| protect | number | No | 0.33 | 0–1 | Safety or backup mechanism for the original audio. |
| main_vocals_volume_change | number | No | 0 | - | Adjustment of the main vocals' volume. |
| backup_vocals_volume_change | number | No | 0 | - | Adjustment of the backup vocals' volume. |
| instrumental_volume_change | number | No | 0 | - | Change in the volume of the instrumental part of the song. |
| pitch_change_all | number | No | 0 | - | Modification of the pitch for all elements of the song. |
| reverb_size | number | No | 0.15 | 0–1 | The perceived size of the reverb effect location. |
| reverb_wetness | number | No | 0.2 | 0–1 | The amount of the reverb effect applied (wet signal). |
| reverb_dryness | number | No | 0.8 | 0–1 | The degree to which the direct sound is present without reverb (dry signal). |
| reverb_damping | number | No | 0.7 | 0–1 | The reduction of high frequencies in the reverb effect. |
| output_format | string | No | mp3 | mp3,wav | The format of the resulting audio file. |

## Example Request

```bash
curl -X POST https://api.eachlabs.ai/v1/prediction/ \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "realistic-voice-cloning",
  "input": {
    "song_input": "https://cdn.eachlabs.ai/ipfs/JsPIizFfRy54Jk5LuXdnrNdV1JHJ6oLmPPdRuIfh3lvpoNai/gangnam.mp3"
  }
}'
```

## Output Schema

Response returned by `GET /v1/prediction/{id}` when the job completes:

```json
{
  "status": "success",
  "predictionID": "string",
  "output": "string (URL of generated audio)",
  "metrics": {
    "predict_time": "number (seconds)"
  }
}
```

## Polling

```bash
curl https://api.eachlabs.ai/v1/prediction/{PREDICTION_ID} \
  -H "Authorization: Bearer YOUR_API_KEY"
```

| Status | Meaning |
|--------|---------|
| `processing` | Still running — poll again |
| `success` | Done — read `output` |
| `error` | Failed — read `message` / `details` |

## Webhook (alternative to polling)

Pass `"webhook_url": "https://your.host/path"` in the create request. Eachlabs POSTs this payload when the job ends:

```json
{
  "exec_id": "prediction-uuid",
  "status": "succeeded",
  "output": "https://...",
  "error": ""
}
```

`status` is `"succeeded"` or `"failed"`. `exec_id` equals the `predictionID` from create. Return 2xx within 30 seconds.

## Errors

Error body: `{ "status": "error", "message": "...", "details": "..." }`

| Code | Meaning |
|------|---------|
| `400` | Invalid input |
| `401` | Missing / invalid `Authorization` bearer token |
| `404` | Unknown model or prediction id |
| `429` | Rate limit — 100 creates / min, 10 concurrent per key |
| `5xx` | Retry with backoff |

## Overview

**realistic-voice-cloning — Voice-to-Voice AI Model**

Developed by eachlabs as part of the eachlabs family, **realistic-voice-cloning** empowers creators to generate song covers and custom audio using any RVC v2 trained AI voice from input audio files, solving the challenge of affordable, high-fidelity voice transformation for music and content production. This **voice-to-voice AI model** excels in cloning realistic voices with minimal input, delivering professional-grade outputs ideal for musicians, podcasters, and developers seeking **realistic voice cloning** tools. Unlike traditional recording sessions, it transforms existing audio into new performances while preserving vocal nuances and style.

## Usage Notes

- API Base URL: `https://api.eachlabs.ai/v1`
- Authentication: send `Authorization: Bearer YOUR_API_KEY`. Generate a key from the Eachlabs dashboard at https://www.eachlabs.ai/dashboard/api-keys.
- File-typed parameters (`*_url`, `image_url`, `video_url`, `audio_url`, etc.) accept publicly-reachable HTTPS URLs only. Upload your asset first (GCS / S3 / your CDN) and pass the resulting URL. Data-URIs and localhost URLs are rejected.
- For structured parameters (arrays / objects) send real JSON values, not stringified payloads.
- Monetary values are reported in USD; per-token / per-megapixel rates may be billed in micro-cents internally.
- Prefer `webhook_url` over polling for long-running predictions — see the Webhook Callback section.