MCP Reference
Complete documentation for Audio Console MCP tools
What is MCP? MCP (Model Context Protocol) is a protocol that allows AI assistants like ChatGPT to interact with external tools and services. Audio Console provides a set of MCP tools for audio editing, conversion, and analysis.
How to use: These tools are automatically available when Audio Console is configured as an MCP server. In ChatGPT and other MCP-compatible clients, you can ask the AI to perform audio operations, and it will use these tools automatically.
UI Tools: Some tools open an interactive UI widget within ChatGPT, allowing you to visually edit audio with waveform visualization, trim controls, and fade effects. Other tools perform operations directly without opening a UI.
Quickstart Guide
Step 1: Get the MCP Server URL
The Audio Console MCP server is available at:
Step 2: Configure Your MCP Client
Add Audio Console to your MCP client configuration. For ChatGPT, add the following to your MCP settings:
{
"mcpServers": {
"audio-console": {
"url": "https://www.audioconsole.app/mcp"
}
}
}
Step 3: Start Using the Tools
Once configured, you can ask ChatGPT or your MCP client to perform audio operations. For example:
- "Open the audio editor with this file"
- "Convert this audio to MP3"
- "Trim the first 30 seconds of this audio"
- "Separate the vocals from this song"
- "Detect the BPM and key of this track"
Step 4: Explore the Tools
Browse the tools below to see all available operations. Tools marked with MCP + UI will open an interactive editor in ChatGPT, while others perform operations directly.
Tools with UI MCP + UI
These tools open an interactive widget within ChatGPT and other MCP clients that support UI rendering. They provide visual waveform editing, trim controls, and real-time preview.
audio.open_audio_editor
Open Audio Editor
Opens an interactive audio editor widget for trimming, fading, and format conversion. Supports MP3, WAV, FLAC, OGG, M4A, and M4R formats. Use this for custom trim points, fade adjustments, or when you want visual waveform editing. The editor provides dual-track processing for vocals and music separation.
Parameters:
audioFile
Object (optional)
Audio file uploaded by the user in the chat. Contains download_url and file_id. Use this when the user attaches an audio file.
audioUrl
String URL (optional)
Public HTTPS URL to an audio file. Use this when the user provides a direct link to an audio file.
audio.open_ringtone_editor
Open Ringtone Editor
Opens a specialized ringtone editor widget optimized for creating ringtones. Same UI as the audio editor but with format options optimized for iPhone (M4R) and Android (OGG). Use this specifically for ringtones. Best for audio clips 30 seconds or less.
Parameters:
audioFile
Object (optional)
Audio file uploaded by the user in the chat. Contains download_url and file_id. Use this when the user attaches an audio file.
audioUrl
String URL (optional)
Public HTTPS URL to an audio file. Use this when the user provides a direct link to an audio file.
Direct Operation Tools
These tools perform operations directly without opening a UI. They process audio and return results immediately. Perfect for automated workflows and batch processing.
audio.convert_from_url
Convert Audio Format
Converts an audio file from a public URL to a different format without editing. Supported formats: MP3, WAV, FLAC, OGG, M4A, M4R. Use this for simple format conversion only. For editing, use audio.open_audio_editor instead.
Parameters:
audioUrl
String URL (required)
Public HTTPS URL of the audio to convert. Example: https://cdn.example.com/audio/song.wav
format
Enum (required)
Target format: mp3, wav, flac, ogg, m4a, m4r
trackName
String (optional, max 80 chars)
Optional display name for the exported file. Example: Session_Mix
audio.trim_start_of_audio
Trim Start of Audio
Extracts exactly the first 30 seconds of an audio file with automatic fade in/out effects. Use this for quick intro extraction. Requires audio to be at least 30 seconds long. For custom trim points, use audio.open_audio_editor instead.
Parameters:
audioUrl
String URL (required)
Public HTTPS URL of the audio file. Example: https://cdn.example.com/audio/song.mp3
format
Enum (optional)
Target audio format. Supported: mp3, wav, flac, ogg, m4a, m4r. Defaults to mp3.
trackName
String (optional, max 80 chars)
Optional display name for the exported file. Example: Intro_30s
audio.trim_end_of_audio
Trim End of Audio
Extracts exactly the last 30 seconds of an audio file with automatic fade in/out effects. Use this for quick outro extraction. Requires audio to be at least 30 seconds long. For custom trim points, use audio.open_audio_editor instead.
Parameters:
audioUrl
String URL (required)
Public HTTPS URL of the audio file. Example: https://cdn.example.com/audio/song.mp3
format
Enum (optional)
Target audio format. Supported: mp3, wav, flac, ogg, m4a, m4r. Defaults to mp3.
trackName
String (optional, max 80 chars)
Optional display name for the exported file. Example: Outro_30s
audio.separate_voice_from_music
Separate Voice from Music
Separates vocals from music and returns both tracks (vocals and instrumental). Use this when you explicitly want both separated tracks. Works best with music that has clear vocal and instrumental separation. Uses LALAL.AI for high-quality separation.
Parameters:
audioFile
Object (optional)
Audio file uploaded by the user in the chat. Contains download_url and file_id.
audioUrl
String URL (optional)
Public HTTPS URL of the audio file to separate. Either audioFile or audioUrl must be provided.
trackName
String (optional, max 80 chars)
Optional display name for the output files. Example: My_Song
audio.remove_vocals
Remove Vocals
Removes vocals from audio and returns only the instrumental/background music track. Use this when you explicitly want instrumental only. Works best with music that has clear vocal and instrumental separation.
Parameters:
audioFile
Object (optional)
Audio file uploaded by the user in the chat. Contains download_url and file_id.
audioUrl
String URL (optional)
Public HTTPS URL of the audio file. Either audioFile or audioUrl must be provided.
trackName
String (optional, max 80 chars)
Optional display name for the output file. Example: My_Song_Instrumental
audio.extract_vocals
Extract Vocals
Extracts vocals from audio and returns only the vocal track without music. Use this when you explicitly want vocals only. Works best with music that has clear vocal and instrumental separation.
Parameters:
audioFile
Object (optional)
Audio file uploaded by the user in the chat. Contains download_url and file_id.
audioUrl
String URL (optional)
Public HTTPS URL of the audio file. Either audioFile or audioUrl must be provided.
trackName
String (optional, max 80 chars)
Optional display name for the output file. Example: My_Song_Vocals
audio.detect_bpm_and_key
Detect BPM and Key
Analyzes audio to detect BPM (beats per minute/tempo) and musical key. Use this for music analysis, DJ mixing, or music production purposes. This tool only provides analysis information - it does not edit or convert audio. Works best with music tracks that have clear rhythm and harmonic content.
Parameters:
audioFile
Object (optional)
Audio file uploaded by the user in the chat. Contains download_url and file_id.
audioUrl
String URL (optional)
Public HTTPS URL of the audio file to analyze. Either audioFile or audioUrl must be provided.
audio.notify_download_link_ready
Share Generated Audio Link
Notifies ChatGPT about audio download links. This tool is typically called automatically by the widget after processing completes. Use this only after the file upload is complete. Primarily used internally by the audio editor widget.
Parameters:
downloadUrl
String URL (required)
Public HTTPS URL where the generated audio can be downloaded. Example: https://downloads.example.com/audio/final.mp3
fileName
String (optional, max 120 chars)
Suggested file name shown to the user. Example: Final.mp3
format
Enum (optional)
Target audio format. Supported: mp3, wav, flac, ogg, m4a, m4r