AllVoiceLab MCP Server

What is AllVoiceLab MCP Server about?

The server provides a set of AI‑driven tools that enable text‑to‑speech, speech‑to‑speech voice conversion, rapid voice cloning, video translation & dubbing, subtitle extraction, subtitle removal, and plain text translation. It is designed to be invoked by MCP‑compatible clients such as Claude Desktop, Cursor, Windsurf, and OpenAI Agents.

How to use AllVoiceLab MCP Server?

Obtain an API key from the AllVoiceLab portal (global or mainland region).

Install the server (recommended via the Python package manager uv):

curl -LsSf https://astral.sh/uv/install.sh | sh
uvx allvoicelab-mcp

Set environment variables required by the server:
- ALLVOICELAB_API_KEY – your personal API key.
- ALLVOICELAB_API_DOMAIN – https://api.allvoicelab.com (global) or https://api.allvoicelab.cn (mainland).
- ALLVOICELAB_BASE_PATH (optional) – directory for storing generated files.
Configure your MCP client (e.g., Claude Desktop) to point to the server using the command and arguments shown above.

Key features of AllVoiceLab MCP Server

Text‑to‑Speech in 30+ languages with ultra‑realistic voices.
Speech‑to‑Speech voice conversion while keeping the original content.
3‑second voice cloning from a short audio sample.
One‑click video translation & dubbing preserving emotional tone and rhythm.
Subtitle extraction via OCR, outputting SRT files.
Hard subtitle removal even on complex backgrounds.
Multilingual text translation with automatic language detection.
Cross‑region API support – choose global or mainland endpoints.

Use cases of AllVoiceLab MCP Server

Audiobook production – generate narrations in multiple languages.
Film & TV localization – translate and dub short videos for global markets.
Game streaming & live content – real‑time voice changing for privacy or character voices.
Content creators – add voice‑overs, replace subtitles, and create multilingual captions.
Enterprise training – produce localized instructional videos quickly.

FAQ from the AllVoiceLab MCP Server

Do I need credits to use the server? Yes, all operations consume AllVoiceLab credits linked to your API key.
Which region should I select? Use the global domain (https://api.allvoicelab.com) for worldwide keys, or the mainland domain (https://api.allvoicelab.cn) for Chinese keys.
What file formats are supported? Text inputs are plain strings; audio files can be MP3, WAV, etc.; video files are MP4, MOV, and common codecs.
Where can I find logs for troubleshooting? Windows: C:\Users\<Username>\.mcp\allvoicelab_mcp.log macOS/Linux: ~/.mcp/allvoicelab_mcp.log
How do I change the output directory? Set the optional environment variable ALLVOICELAB_BASE_PATH to the desired folder path.
Can I run multiple instances simultaneously? Yes, each instance can use a separate base path and distinct API keys.

Official AllVoiceLab Model Context Protocol (MCP) server, supporting interaction with powerful text-to-speech and video translation APIs. Enables MCP clients like Claude Desktop, Cursor, Windsurf, OpenAI Agents to generate speech, translate videos, and perform intelligent voice conversion. Serves scenarios such as short drama localization for global markets, AI-Generated audiobooks, AI-Powered production of film/TV narration.

Why Choose AllVoiceLab MCP Server?

Multi-engine technology unlocks infinite possibilities for voice: With simple text input, you can access video generation, speech synthesis, voice cloning, and more.
AI Voice Generator (TTS): Natural voice generation in 30+ languages with ultra-high realism
Voice Changer: Real-time voice conversion, ideal for gaming, live streaming, and privacy protection
Multilingual Dubbing: One-click translation and dubbing for short videos/films, preserving emotional tone and rhythm
Speech-to-Text (STT): AI-powered multilingual subtitle generation with over 98% accuracy
Subtitle Removal: Seamless hard subtitle erasure, even on complex backgrounds
Voice Cloning: 3-Second Ultra-Fast Cloning with Human-like Voice Synthesis

Documentation

中文文档

Quickstart

Get your API key from AllVoiceLab.
Install uv (Python package manager), install with curl -LsSf https://astral.sh/uv/install.sh | sh
Important: The server addresses of APIs in different regions need to match the keys of the corresponding regions, otherwise there will be an error that the tool is unavailable.

Region	Global	Mainland
ALLVOICELAB_API_KEY	go get from AllVoiceLab	go get from AllVoiceLab
ALLVOICELAB_API_DOMAIN	https://api.allvoicelab.com	https://api.allvoicelab.cn

Claude Desktop

Go to Claude > Settings > Developer > Edit Config > claude_desktop_config.json to include the following:

{
  "mcpServers": {
    "AllVoiceLab": {
      "command": "uvx",
      "args": ["allvoicelab-mcp"],
      "env": {
        "ALLVOICELAB_API_KEY": "<insert-your-api-key-here>",
        "ALLVOICELAB_API_DOMAIN": "<insert-api-domain-here>",
        "ALLVOICELAB_BASE_PATH":"optional, default is user home directory.This is uesd to store the output files."
      }
    }
  }
}

If you're using Windows, you will have to enable "Developer Mode" in Claude Desktop to use the MCP server. Click "Help" in the hamburger menu in the top left and select "Enable Developer Mode".

Cursor

Go to Cursor -> Preferences -> Cursor Settings -> MCP -> Add new global MCP Server to add above config.

That's it. Your MCP client can now interact with AllVoiceLab.

Available methods

Methods	Brief description
text_to_speech	Convert text to speech
speech_to_speech	Convert audio to another voice while preserving the speech content
clone_voice	Create a custom voice profile by cloning from an audio sample
remove_subtitle	Remove hardcoded subtitles from a video using OCR
video_translation_dubbing	Translate and dub video speech into different languages
text_translation	Translate a text file into another language
subtitle_extraction	Extract subtitles from a video using OCR

Example usage

⚠️ Warning: AllVoiceLab credits are needed to use these tools.

1. Text to Speech

Try asking: Convert "At All Voice Lab, we’re reshaping the future of audio workflows with AI-powered solutions, making authentic voices accessible to creators everywhere." into voice.

2. Voice Conversion

After generating the audio from the previous example, select the audio file and ask: Convert this to a male voice.

3. Voice Cloning

Select an audio file with a single voice and ask: Clone this voice.

4. Video Translation

Select a video file (English) and ask: Translate this video to japanese.

Original video:

After translation:

5. Remove Subtitles

Select a video with subtitles and ask: Remove the subtitles from this video.

Original video:

After the task is completed:

6. Text Translation

Select a long text (for example, "The Foolish Old Man Removes the Mountains") and ask: Translate this text to japanese. If no language is specified, it will be translated to English by default.

7. Subtitle Extraction

Select a video with subtitles and ask: Extract the subtitles from this video.

After the task is completed, you will get an SRT file as shown below:

Troubleshooting

Logs can be found at:

Windows: C:\Users<Username>.mcp\allvoicelab_mcp.log
macOS: ~/.mcp/allvoicelab_mcp.log

Please contact us by email(tech@allvoicelab.com) with log files