pyVideoTrans

Appearance

TTS Channel: OmniVoice-TTS ​

What is OmniVoice ​

OmniVoice is an advanced zero-shot multilingual TTS model with the following features:

  • Extensive language support: 600+ languages
  • High-quality speech: Natural, fluent output
  • Voice cloning: Clone voices from reference audio
  • Fast inference: Excellent inference speed

pyVideoTrans has supported OmniVoice since v3.98-0403.

OmniVoice open-source: https://github.com/k2-fsa/OmniVoice

Prerequisites ​

RequirementDetails
pyVideoTrans version≥ v3.98-0403
Operating SystemWindows (package), Linux/macOS (source deployment)
HardwareNVIDIA GPU recommended
NetworkFirst launch downloads models from huggingface.co

Download ​

Start ​

  1. Download the .7z file and extract it
  2. Double-click start.bat
  3. Wait for startup to complete — the success indicator is when the following screen appears and stays:

The first launch will download models, defaulting to the China mirror at https://hf-mirror.com.

MacOS/Linux Source Deployment ​

Step 1: Download Source Code ​

  1. Go to the OmniVoice official repository and download the source code
  2. Click the green <>Code button in the middle of the homepage, then click Download ZIP
  3. After extraction, navigate to the folder containing pyproject.toml (usually inside OmniVoice-master)

Step 2: Install uv ​

Open a terminal and run:

bash
curl -LsSf https://astral.sh/uv/install.sh | sh

Windows users: Download the uv archive, extract it, and place uv.exe alongside pyproject.toml.

Step 3: Install Dependencies ​

Open a terminal in the folder containing pyproject.toml and run:

bash
uv sync

Step 4: Configure Model Download Mirror ​

The model hosting site huggingface.co may not be accessible directly. Set a mirror:

Windows:

bash
set HF_ENDPOINT=https://hf-mirror.com

MacOS/Linux:

bash
export HF_ENDPOINT=https://hf-mirror.com

Step 5: Start the Server ​

bash
uv run omnivoice-demo --ip 0.0.0.0 --port 8081

Using in pyVideoTrans ​

Configuration Steps ​

  1. Open pyVideoTrans
  2. Go to Menu → TTS Settings → OmniVoice-URL
  3. Enter the address: http://127.0.0.1:8081
  4. Fill in the reference audio and corresponding text in the character list below

Reference Audio Format ​

audio_filename.wav#text corresponding to the audio

Example: If you have an audio file nverguo.wav (a clean recording of the Queen of Women speaking), place it in the f5-tts folder under pyVideoTrans, then enter in the settings:

nverguo.wav#Enter the speech text here

  1. Click "Test" — if it shows OK, click Save, and you're ready to use it

Using on Google Colab ​

You can also use it directly on Colab by opening this link:

  1. Click "Run All" at the top
  2. Wait for a URL like https://xxxxxxxx.gradio.live to appear in the bottom cell
  3. Click to open it in your browser
  4. Copy this address to pyVideoTrans Menu → TTS Settings → OmniVoice-URL

Reference Audio Requirements ​

ItemRequirement
FormatWAV (recommended)
Duration3–10 seconds (ideal)
ContentClear pronunciation, no background noise
Locationf5-tts folder under pyVideoTrans root directory

Source code verification (_omnivoice.py): OmniVoice uses the /_clone_fn Gradio API endpoint, supports 60+ language code mappings, and has a built-in retry mechanism (tenacity). It also supports OmniVoice-Studio (port 3900) using the REST API /generate endpoint.

Troubleshooting ​

1. Can't see pyproject.toml after source deployment? ​

After extraction, you'll see an OmniVoice-master folder. Open it to find pyproject.toml or pyproject — all operations should be done inside this folder.

2. Error: uv is not recognized as an internal or external command? ​

uv is not installed or not in your system PATH. Windows users: download the uv archive, extract it, and copy uv.exe to the same directory as pyproject.toml. Other systems: run the install command directly.

3. Model download is slow or fails? ​

Models are downloaded from huggingface.co, which may not be accessible directly. Set the model mirror:

Windows:

bash
set HF_ENDPOINT=https://hf-mirror.com

MacOS/Linux:

bash
export HF_ENDPOINT=https://hf-mirror.com

4. uv sync fails? ​

Module installation defaults to downloading from overseas sources. Use the Alibaba mirror:

bash
uv sync --default-index "https://mirrors.aliyun.com/pypi/simple"

5. Still failing with the Alibaba mirror? ​

You may be missing the MSVC environment. Install it following this guide.

6. How to use OmniVoice directly in the browser? ​

After startup, open http://127.0.0.1:8081 in your browser.

7. How to change the startup port and model mirror in the package? ​

Open start.bat with a text editor:

  • To disable the China mirror, remove the line set "HF_ENDPOINT=https://hf-mirror.com"
  • To change the port, modify the 8081 in set "PORT=8081" to your desired port