Zsxkib Voice Train
Model Information
Display Name: Zsxkib Voice Train
API Model ID: zsxkib/train
Category: Voice Training
Description: Train a custom RVC v2 voice model from your audio recordings. Once trained, use the voice_id with zsxkib/voice to generate AI voice covers of any song. ## How It Works 1. Prepare a zip file with clean voice recordings (10-30 min ideal, singing preferred) 2. Upload via the audio parameter 3. Receive a voice_id for use with zsxkib/voice ## Parameters | Parameter | Type | Required | Default | Description | |-----------|------|----------|---------|-------------| | audio | URL | Yes | — | Zip file with voice recordings (WAV format, 10-30 min total) | | sample_rate | string | No | 48k | Audio sample rate: 40k or 48k | | version | string | No | v2 | RVC version: v1 or v2 | | f0method | string | No | rmvpe_gpu | Pitch extraction: pm, dio, harvest, rmvpe, rmvpe_gpu | | epoch | number | No | 10 | Training epochs (more = better quality, slower) | | batch_size | string | No | 7 | Training batch size | ## Output Returns JSON with voice_id (for use with zsxkib/voice).
How to Use This Model
To use Zsxkib Voice Train via the HInow.ai API, use the model ID: zsxkib/train
API Request Example
POST https://api.hinow.ai/v1/chat/completions
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json
{
"model": "zsxkib/train",
"messages": [
{"role": "user", "content": "Your input here"}
]
}
Pricing
- total: $0.50
Available Parameters
- audio: Zip file URL with voice recordings (WAV format, 10-30 min of clean audio)
- sample_rate: Audio sample rate. Default: 48k (Options: 48k, 40k)
- version: RVC version. Default: v2 (Options: v2, v1)
- f0method: Pitch extraction method. Default: rmvpe_gpu (Options: rmvpe_gpu, rmvpe, harvest, dio, pm)
- epoch: Training epochs (more = better quality, slower). Default: 10
- batch_size: Training batch size. Default: 7
Quick Reference
To use this model, set: "model": "zsxkib/train"
Featured: No
Documentation: https://hinow.ai/models/zsxkib/train
API Endpoint: https://api.hinow.ai/v1
