Model Information

Display Name: Pseudoram RVC v2

API Model ID: pseudoram/rvc-v2

Category: Voice Cover

Description: RVC v2 voice conversion model. Converts any audio (speech, vocals) to a trained voice. Faster and cheaper than zsxkib/voice, but does not separate vocals from instrumentals — best for isolated voice audio or speech-to-speech conversion. ## How It Works 1. Train your voice first using zsxkib/train (one-time) 2. Send any audio file + your voice_id 3. Get back the audio with your trained voice ## Difference from zsxkib/voice - **pseudoram/rvc-v2**: Converts audio directly. Best for speech or isolated vocals. Faster (~55s) and cheaper ($0.02/run). - **zsxkib/voice**: Full song pipeline — separates vocals/instrumentals, converts voice, remixes. Best for complete songs. Slower (~3min) at $0.10/run.

How to Use This Model

To use Pseudoram RVC v2 via the HInow.ai API, use the model ID: pseudoram/rvc-v2

API Request Example

POST https://api.hinow.ai/v1/chat/completions
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
  "model": "pseudoram/rvc-v2",
  "messages": [
    {"role": "user", "content": "Your input here"}
  ]
}
              

Pricing

• total: $0.02

Available Parameters

• audio: Audio file URL to convert (MP3, WAV). Speech or isolated vocals work best.
• voice_id: Voice ID from zsxkib/train
• pitch_change: Pitch shift in semitones. Positive = higher, negative = lower. Default: 0
• index_rate: AI accent retention 0-1 (0=less, 1=more). Default: 0.5
• filter_radius: Audio smoothing 0-7. Default: 3
• rms_mix_rate: Original loudness mix 0-1. Default: 0.25
• f0_method: Pitch detection. rmvpe=clarity, mangio-crepe=smoother. Default: rmvpe (Options: rmvpe, mangio-crepe)
• crepe_hop_length: Pitch check frequency in ms (only for mangio-crepe). Default: 128
• protect: Preserve breath sounds 0-0.5. Default: 0.33
• output_format: Output audio format. Default: mp3 (Options: mp3, wav)

Quick Reference

To use this model, set: "model": "pseudoram/rvc-v2"

Featured: No

Documentation: https://hinow.ai/models/pseudoram/rvc-v2

API Endpoint: https://api.hinow.ai/v1