# ExpertVagabond/rvc-mcp

Source: https://glimind.com/server/github%2Fexpertvagabond%2Frvc-mcp
Registry id: `github/expertvagabond/rvc-mcp`
Status: unknown

## Tools (12)

- `rvc_clean`: Unload voice models from memory (free RAM/VRAM)
- `rvc_export_onnx`: Convert model to ONNX format for cross-platform inference
- `rvc_extract_features`: Step 2: Extract F0 pitch curves and HuBERT features
- `rvc_infer`: Convert vocals to a target voice (all F0 methods, index support, pitch shift)
- `rvc_list_models`: List available .pth voice models and .index files
- `rvc_model_extract`: Extract portable model from large training checkpoint
- `rvc_model_info`: View model metadata (architecture, sample rate, training info)
- `rvc_model_merge`: Blend two voice models by weighted average
- `rvc_preprocess`: Step 1: Slice audio into segments, normalize, resample
- `rvc_separate_vocals`: Separate vocals from accompaniment using UVR5 (HP2/HP3/HP5/MDX-Net/DeEcho)
- `rvc_status`: Check if RVC WebUI is running, get start command
- `rvc_train`: Step 3: Train RVC model and build FAISS index (30-60+ min)

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.