transcribe_audio

Convert speech in audio files to text for transcription, supporting multiple languages and optional webhook notifications.

Instructions

Transcribe speech from audio to text

Input Schema

TableJSON Schema

Name	Required	Description
`audio_url`	Yes	URL of the audio file to transcribe
`language`	No	Language code (e.g., 'en', 'es', 'fr')
`webhook_url`	No	URL for callback upon completion

Implementation Reference

src/index.ts:1209-1228 (handler)

The handler function that implements the core logic of the transcribe_audio tool by making an API call to /audiotranscribe endpoint.

private async handleTranscribeAudio(args: any) {
  if (!args.audio_url) {
    throw new McpError(ErrorCode.InvalidParams, "audio_url is required");
  }

  const response = await this.axiosInstance.post("/audiotranscribe", {
    audio_url: args.audio_url,
    language: args.language,
    webhook_url: args.webhook_url,
  });

  return {
    content: [
      {
        type: "text",
        text: `Audio transcription started!\n\n${JSON.stringify(response.data, null, 2)}\n\nUse get_conversion_by_id with the task_id to check the status.`,
      },
    ],
  };
}

src/index.ts:524-541 (schema)

Input schema defining parameters for the transcribe_audio tool including audio_url (required), language, and webhook_url.

inputSchema: {
  type: "object" as const,
  properties: {
    audio_url: {
      type: "string",
      description: "URL of the audio file to transcribe",
    },
    language: {
      type: "string",
      description: "Language code (e.g., 'en', 'es', 'fr')",
    },
    webhook_url: {
      type: "string",
      description: "URL for callback upon completion",
    },
  },
  required: ["audio_url"],
},

src/index.ts:522-542 (registration)

Tool registration object in the TOOLS array that defines the tool's metadata and schema for MCP list tools request.

  name: "transcribe_audio",
  description: "Transcribe speech from audio to text",
  inputSchema: {
    type: "object" as const,
    properties: {
      audio_url: {
        type: "string",
        description: "URL of the audio file to transcribe",
      },
      language: {
        type: "string",
        description: "Language code (e.g., 'en', 'es', 'fr')",
      },
      webhook_url: {
        type: "string",
        description: "URL for callback upon completion",
      },
    },
    required: ["audio_url"],
  },
},

src/index.ts:713-714 (registration)
Switch case in the CallTool request handler that registers and routes transcribe_audio calls to its handler function.
```
case "transcribe_audio":
  return await this.handleTranscribeAudio(args);
```

MusicGPT MCP Server