SunoMCP

Instructions

This saves the vocal characteristics from a generated song so you can reuse
that same voice style in future generations. Great for maintaining consistency
across multiple songs.

Use this when:
- You generated a song and love the voice
- You want to create multiple songs with the same vocalist
- You're building an album with consistent vocal style
- You want to save a unique voice for future use

After creating a persona, use suno_generate_with_persona with the returned
persona_id to generate new songs with that voice.

Returns:
    Persona ID that can be used with suno_generate_with_persona tool.

Implementation Reference

• tools/persona_tools.py:12-82 (handler)

  Main tool handler function for suno_create_persona. Decorated with @mcp.tool(), it accepts audio_id, name, and optional parameters (vox_audio_id, vocal_start, vocal_end, description) to create a reusable vocal persona from an existing audio. Calls client.create_persona() and returns formatted result.

  @mcp.tool()
  async def suno_create_persona(
      audio_id: Annotated[
          str,
          Field(
              description="ID of the audio to use as the persona reference. This should be a previously generated song whose vocal style you want to save and reuse."
          ),
      ],
      name: Annotated[
          str,
          Field(
              description="Name for this persona. Use a descriptive name that helps you remember the voice style. Examples: 'My Rock Voice', 'Soft Female Singer', 'Deep Male Baritone', 'Energetic Pop Vocalist'"
          ),
      ],
      vox_audio_id: Annotated[
          str | None,
          Field(
              description="Optional audio ID used to generate a new singer's style by combining with the main audio. Useful for creating hybrid vocal personas."
          ),
      ] = None,
      vocal_start: Annotated[
          float | None,
          Field(
              description="Start time in seconds of the vocal segment to use from the audio. Useful for isolating a specific vocal section."
          ),
      ] = None,
      vocal_end: Annotated[
          float | None,
          Field(description="End time in seconds of the vocal segment to use from the audio."),
      ] = None,
      description: Annotated[
          str | None,
          Field(
              description="Description of the singer's style. Examples: 'Warm and breathy female voice with jazz influences', 'Powerful male rock vocalist with raspy tone'"
          ),
      ] = None,
  ) -> str:
      """Create a new artist persona from an existing audio's vocal style.
  
      This saves the vocal characteristics from a generated song so you can reuse
      that same voice style in future generations. Great for maintaining consistency
      across multiple songs.
  
      Use this when:
      - You generated a song and love the voice
      - You want to create multiple songs with the same vocalist
      - You're building an album with consistent vocal style
      - You want to save a unique voice for future use
  
      After creating a persona, use suno_generate_with_persona with the returned
      persona_id to generate new songs with that voice.
  
      Returns:
          Persona ID that can be used with suno_generate_with_persona tool.
      """
      payload: dict = {
          "audio_id": audio_id,
          "name": name,
      }
  
      if vox_audio_id:
          payload["vox_audio_id"] = vox_audio_id
      if vocal_start is not None:
          payload["vocal_start"] = vocal_start
      if vocal_end is not None:
          payload["vocal_end"] = vocal_end
      if description:
          payload["description"] = description
  
      result = await client.create_persona(**payload)
      return format_persona_result(result)

• core/utils.py:91-104 (helper)

  Helper function format_persona_result that formats the persona creation API response as a JSON string with submission guidance for follow-up task queries.

  def format_persona_result(data: dict[str, Any]) -> str:
      """Format persona creation result as JSON.
  
      Args:
          data: API response dictionary
  
      Returns:
          JSON string representation of the result
      """
      return json.dumps(
          _with_submission_guidance(data, "suno_get_task", "suno_get_tasks_batch"),
          ensure_ascii=False,
          indent=2,
      )

• main.py:239-241 (registration)

  Registration of suno_create_persona tool in the main capabilities list with description 'Save a voice style for reuse'. This is part of the tools listing sent to the MCP client.

      "name": "suno_create_persona",
      "description": "Save a voice style for reuse",
  },