Gaudio Developers MCP

What gaudiolab-jp/gaudio-developers-mcp MCP server does

The gaudiolab-jp/gaudio-developers-mcp MCP server connects MCP-compatible clients to the Gaudio Lab Audio AI API. It supports three main workflows: separating musical stems, separating dialogue/music/effects tracks, and aligning lyrics with timestamps.

Stem separation models can produce vocal, drum, bass, electric guitar, acoustic piano, or speech outputs depending on the selected model. DME models handle dialogue, vocals, music, and effects in different combinations. The lyrics workflow accepts UTF-8 text files and returns synchronized lyric data as CSV with timestamps, lyric text, and confidence scores, along with a JSON report.

Supported inputs include WAV, FLAC, MP3, M4A, MOV, MP4, and UTF-8 TXT files. Video inputs have their audio extracted automatically. Processing outputs include MP3 at 48 kHz and 320 kbps, plus WAV in the same format as the input where applicable. Download URLs remain valid for 48 hours.

How it works

The gaudiolab-jp/gaudio-developers-mcp MCP server can perform the API workflow in stages: upload a file, create a processing job, poll the job status, and retrieve download URLs. Upload identifiers remain reusable for 72 hours, allowing multiple jobs to use the same uploaded file.

Two higher-level operations wrap this sequence. gaudio_separate_audio handles upload, processing, and result retrieval for audio separation. gaudio_sync_lyrics provides the corresponding end-to-end workflow for lyric synchronization. The staged tools are useful when an agent needs to inspect or control individual steps.

File limits depend on the model family. Stem separation accepts files up to 1 GB and 20 minutes, DME separation accepts up to 10 GB and 200 minutes, and lyric synchronization accepts up to 1 GB and 10 minutes. Lyric files must contain at least two lines, with a maximum of 60 characters per line.

Setup and configuration

The gaudiolab-jp/gaudio-developers-mcp MCP server requires a Gaudio Developers account, a project, and an API key created in the Gaudio dashboard. Set the key in the GAUDIO_API_KEY environment variable.

A local stdio configuration runs the published npm package with npx:

{
  "mcpServers": {
    "gaudio": {
      "command": "npx",
      "args": ["-y", "@gaudiolab/mcp-developers"],
      "env": {
        "GAUDIO_API_KEY": "your-api-key-here"
      }
    }
  }
}

The package is intended for MCP-compatible clients, including Claude, Cursor, VS Code, GitHub Copilot, and other clients that support MCP. The API key controls project access, available models, and credit information.

Tools and capabilities

• gaudio_get_key_info reports credits, project details, and permitted models.
• gaudio_list_models lists available models by category.
• gaudio_upload_file uploads audio, video, or text using multipart transfer with automatic chunking.
• gaudio_create_job starts a processing job.
• gaudio_get_job checks job status and returns download URLs.
• gaudio_separate_audio combines upload, processing, and result retrieval for separation tasks.
• gaudio_sync_lyrics runs lyric synchronization and returns timestamped results.

The repository is licensed under MIT. API access requires the Gaudio Developers service and an API key; the provided material does not specify service pricing or credit costs.