pyVideoTrans

Appearance

TTS Service: Spark-TTS / Index-TTS / Dia-TTS / VoxCPM / Confucius-TTS ​

Since v3.68, these TTS services share the same settings interface. Simply enter the WebUI URL (default: http://127.0.0.1:7860) and select the corresponding service from the "TTS Service" dropdown on the main interface.

This document covers the deployment and integration of 6 open-source TTS services, all accessible via Gradio WebUI API with largely identical configuration.

1. Service Overview ​

ServiceDeveloperLanguagesDefault URLOfficial Project
Spark-TTSSparkAudioChinese, Englishhttp://127.0.0.1:7860https://github.com/SparkAudio/Spark-TTS
Index-TTSBilibili Open SourceChinese, Englishhttp://127.0.0.1:7860https://github.com/index-tts/index-tts
VoxCPM-TTSOpenBMB30+ languageshttp://127.0.0.1:7860https://github.com/OpenBMB/VoxCPM
Confucius-TTS (built-in since v4.06)NetEase Youdao14 languageshttp://127.0.0.1:7860https://github.com/netease-youdao/Confucius4-TTS

All services follow the same integration flow: start WebUI → enter URL in pyVideoTrans → select service → configure reference audio → use.

2. Prerequisites ​

  1. pyVideoTrans version: v3.68 or later
  2. Hardware: NVIDIA GPU recommended (GPU acceleration). Some services support CPU (slower).
  3. Network: First launch may download models from HuggingFace. A VPN or mirror may be needed in China.
  4. Python: For source deployment, Python 3.9+ and Git are required.

Index-TTS ​

What is Index-TTS ​

Index-TTS is an open-source TTS service from Bilibili, supporting Chinese and English voice cloning with excellent audio quality.

Deployment ​

Windows Package ​

Source Deployment ​

Follow the official documentation: https://github.com/index-tts/index-tts

Launching WebUI ​

The default address is http://127.0.0.1:7860. Configuration is identical to F5-TTS.

Note: Only Index-TTS official WebUI is supported. If using a third-party package, replace its webui.py with the official one, or it may not work.

VoxCPM-TTS ​

What is VoxCPM-TTS ​

VoxCPM-TTS is an open-source TTS service from OpenBMB, supporting 30+ languages including multiple Chinese dialects.

Supported Languages ​

  • International (30+): Arabic, Burmese, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Indonesian, Italian, Japanese, Khmer, Korean, Lao, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Tagalog, Thai, Turkish, Vietnamese
  • Chinese dialects: Mandarin, Sichuan, Cantonese, Wu, Northeastern, Henan, Shaanxi, Shandong, Tianjin, Southern Min

Deployment ​

Source deployment is required. Follow the official documentation: https://github.com/OpenBMB/VoxCPM

The default WebUI address is http://127.0.0.1:7860. Configuration is identical to F5-TTS.

Spark-TTS ​

What is Spark-TTS ​

Spark-TTS is an open-source TTS service from SparkAudio, supporting Chinese and English voice cloning.

Deployment ​

Source deployment is required. Follow the official documentation: https://github.com/SparkAudio/Spark-TTS

The default WebUI address is http://127.0.0.1:7860. Configuration is identical to F5-TTS.

Confucius-TTS [Built-in since v4.06] ​

What is Confucius-TTS ​

Confucius-TTS is an open-source TTS service from NetEase Youdao, supporting 14 languages.

Supported Languages ​

Chinese, English, Japanese, Korean, German, French, Spanish, Indonesian, Italian, Thai, Portuguese, Russian, Malay, Vietnamese

Deployment ​

Windows Package ​

Double-click start-webui.bat to use.

Source Deployment (Extra Steps Needed) ​

The official source doesn't include WebUI or API. After source deployment:

  1. Download webui.py
  2. Place webui.py in the official source directory
  3. Install gradio: pip install gradio
  4. Launch: python webui.py

Default address: http://127.0.0.1:7860. Configuration is identical to F5-TTS.

Configuring Reference Audio ​

Reference audio is configured uniformly via Menu → TTS Settings → Set Reference Audio.

Steps ​

  1. Open the "Reference Audio" settings interface.
  2. Enter content in the following format:
    audio_filename.wav#text spoken in the audio file
  3. Place the reference audio file in the f5-tts folder under the pyVideoTrans root directory (create it if it doesn't exist).

Example ​

If you have an audio file nverguo.wav containing "女儿国王说话", enter:

nverguo.wav#女儿国王说话

Place reference audio in pyVideoTrans's f5-tts folder

Reference audio and its text content

Reference Audio Requirements ​

ItemRequirement
FormatWAV (recommended), MP3 also acceptable
Duration3~12 seconds (F5-TTS max 12s, longer auto-truncated)
ContentClear pronunciation, no background noise
TextMust match the audio content

Common Errors & Notes ​

1. Don't close the terminal window ​

While using the API, you can close the browser WebUI, but do NOT close the terminal window that launched the TTS service.

This window must remain open, or the API won't work

2. Cannot switch models dynamically ​

Model switching is not supported dynamically. Modify the code and restart the WebUI.

3. HuggingFace connection timeout ​

If you see:

requests.exceptions.ConnectTimeout: HTTPSConnectionPool(host='huggingface.co', port=443): Max retries exceeded

This is a network issue. Configure a VPN or proxy to access huggingface.co.

4. API Endpoints ​

ServiceGradio API Endpoint
Spark-TTS/voice_clone
Index-TTS/gen_single
VoxCPM-TTS/generate
Confucius-TTS/_clone_fn

These endpoints are called internally by the software — no manual configuration needed.

5. Reference Audio Duration ​

Recommended 3~10 seconds.

6. Version Compatibility ​

  • Ensure pyVideoTrans ≥ v3.68
  • Use official WebUI for each TTS service (third-party packages may have incompatible APIs)