### Install Dependencies from requirements.txt

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Install project dependencies by running this command after cloning the repository. Ensure Python and PyTorch are already installed.

```bash
pip install -r requirements.txt
```

--------------------------------

### Install GPU Extra for Audio Separator

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the audio-separator library with the [gpu] extra to enable GPU support. This is the primary method for GPU installation.

```sh
pip install audio-separator[gpu]
```

--------------------------------

### Install Dependencies and Modal CLI

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/audio_separator/remote/README.md

Install necessary project dependencies and the Modal CLI for deploying the API server. Ensure you are in a virtual environment with audio-separator dependencies installed.

```bash
poetry install --extras cpu
pip install modal
modal setup
```

--------------------------------

### Install Demucs from Git Repository

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Install Demucs by downloading the git repository and running the pip install command within PowerShell. Ensure you have the correct directory path.

```bash
pip install *insert the directory of the extracted repo here*
```

--------------------------------

### Install Audio Separator with Pip (CPU)

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the package for CPU-only usage. Ensure FFmpeg is installed separately.

```bash
pip install audio-separator
```

--------------------------------

### Install PyTorch

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Install a specific version of PyTorch to resolve potential unpickling issues. This is a prerequisite for running some audio separation models.

```python
!pip install torch==2.5.0
```

--------------------------------

### Install PyTorch from local wheels

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Install PyTorch from local wheel files by providing the full path to the .whl file. Use quotes if the path contains spaces.

```bash
pip install SomePackage-1.0-py2.py3-none-any.whl
```

--------------------------------

### Install PyTorch with ROCm 6.3

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Installs PyTorch and torchvision for ROCm 6.3. This is an alternative for users with ROCm 6.x installations.

```bash
torch 2.9.0+rocm6.3 torchvision0.24.0+rocm6.3 --index-url https://download.pytorch.org/whl/rocm6.3
```

--------------------------------

### Example: Separate with Multiple Models and Custom Settings

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/audio_separator/remote/README.md

Demonstrates separating audio using multiple specified models, custom output format, normalization, API URL, and timeout settings.

```bash
# Separate with multiple models and custom settings
audio-separator-remote separate song.mp3 \
  --models model_bs_roformer_ep_317_sdr_12.9755.ckpt UVR-MDX-NET-Inst_HQ_4.onnx \
  --output_format wav \
  --normalization 0.8 \
  --api_url https://my-api.com \
  --timeout 600
```

--------------------------------

### Install FFmpeg on macOS

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the FFmpeg package on macOS using Homebrew. This command updates Homebrew and then installs FFmpeg.

```sh
brew update; brew install ffmpeg
```

--------------------------------

### Install GPU Dependencies with Poetry

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install extra dependencies required for GPU acceleration using Poetry. This is an alternative to CPU-only installation.

```shell
poetry install --extras "gpu"
```

--------------------------------

### Local Installation with inference.py

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Use this script for local inference if you have cloned the repository. The command line can be extensive.

```python
python inference.py \
--input_audio "{file_path}" \
--large_gpu \
--BSRoformer_model {BSRoformer_model} \
--weight_BSRoformer {weight_BSRoformer} \
--weight_Kim_MelRoformer {weight_Kim_MelRoformer} \
--weight_InstVoc {weight_InstVoc} \
--weight_InstHQ4 {weight_InstHQ4} \
--weight_VOCFT {weight_VOCFT} \
--weight_VitLarge {weight_VitLarge} \

```

--------------------------------

### Install ONNX Runtime GPU (Alternative)

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

An alternative installation command for ONNX Runtime GPU, useful if you encounter credential issues with the nightly build.

```bash
!python -m pip -q install onnxruntime-gpu --extra-index-url=https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/
```

--------------------------------

### Install Audio Separator (CPU Only - Conda)

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the audio-separator package using Conda for CPU-only inference. This is a minimal installation without GPU dependencies.

```sh
conda install audio-separator -c pytorch -c conda-forge
```

--------------------------------

### Install Audio Separator with Pip (Nvidia GPU)

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the package with support for Nvidia GPUs using CUDA. This requires a compatible CUDA toolkit installation.

```bash
pip install audio-separator[cuda]
```

--------------------------------

### Example: Separate with Custom Output Names

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/audio_separator/remote/README.md

Shows how to specify custom names for the output stems during audio separation.

```bash
# Separate with custom output names
audio-separator-remote separate song.mp3 \
  --model htdemucs_6s.yaml \
  --custom_output_names '{"Vocals": "vocals", "Drums": "drums", "Bass": "bass", "Other": "other"}'
```

--------------------------------

### Install Audio Separator with GPU Support (Pip)

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the audio-separator package with GPU acceleration using pip. This command installs the necessary dependencies for GPU inference.

```sh
pip install "audio-separator[gpu]"
```

--------------------------------

### Install CPU Dependencies with Poetry

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install extra dependencies specifically for CPU execution using Poetry. Use this if GPU support is not needed or available.

```shell
poetry install --extras "cpu"
```

--------------------------------

### List Available Models via CLI

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Use the command-line interface to list all available audio separation models. This helps in choosing the right model for your task.

```bash
audio-separator --list-models
```

--------------------------------

### Install samplerate for Demucs Module Fix

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Use this command to install a specific version of samplerate before installing requirements to resolve Demucs module not found errors in Colab environments.

```bash
!pip install samplerate==0.1.0
!pip install -r requirements.txt &> /dev/null
```

--------------------------------

### Install Missing Module

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

If you encounter a ModuleNotFoundError for 'hyper_connections', install it using pip.

```bash
pip install hyper_connections
```

--------------------------------

### Install Intel C++ Compiler Runtime

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Use this command to resolve LLVM errors when setting up the MDX23 music separation model.

```bash
conda install -c intel icc_rt
```

--------------------------------

### Install PyTorch with CUDA 10 using --extra-index-url

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/docs/deton24-audio-separation-info-2026-03-15.md

Use this command to install PyTorch versions compatible with CUDA 10 when direct installation fails. This is often necessary for older GPUs.

```bash
pip install torch==2.3.0 --extra-index-url https://download.pytorch.org/whl/cu100
```

--------------------------------

### Install Project Dependencies with Poetry

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/README.md

Install the project's core dependencies using Poetry. This command ensures all necessary libraries are available for the project.

```shell
poetry install
```

--------------------------------

### Example: Use VR Parameters for Vocal Isolation

Source: https://github.com/nomadkaraoke/python-audio-separator/blob/main/audio_separator/remote/README.md

Shows how to use specific VR (Vocal Isolation) parameters to improve the isolation of vocals from an audio file.

```bash
# Use VR parameters for better vocal isolation
audio-separator-remote separate vocals.wav \
  --model 2_HP-UVR.pth \
  --vr_aggression 15 \
  --vr_window_size 320 \
  --vr_enable_tta
```