Generate Musical Masterpieces with Charles McCarthy's MusicGen Cognitive Actions

24 Apr 2025
Generate Musical Masterpieces with Charles McCarthy's MusicGen Cognitive Actions

In the realm of music generation, the Charles McCarthy MusicGen API stands out by providing developers with powerful tools to create high-quality music tracks effortlessly. The Cognitive Actions within this API allow you to generate music of up to 60 seconds, employing advanced models that balance cost and quality. Whether you're building an application for music composition, sound design, or simply exploring creative audio applications, these pre-built actions can significantly streamline your development process.

Prerequisites

Before diving into the MusicGen Cognitive Actions, ensure you have the following:

  • An API key for the Cognitive Actions platform, which will be needed for authentication.
  • Basic knowledge of JSON and HTTP requests, as you'll be interacting with a web API.
  • Familiarity with Python, as we'll provide a conceptual example using Python to illustrate how to invoke these actions.

To authenticate with the API, you'll typically include your API key in the headers of your HTTP requests.

Cognitive Actions Overview

Generate Music with MusicGen

The Generate Music with MusicGen action is designed to create music pieces based on textual prompts or existing audio files. This action excels in generating music efficiently, optimizing for performance and cost without sacrificing quality.

Input

The input schema requires the following fields, with several optional parameters to fine-tune the output:

  • seed (integer): A random seed for generation; if set to -1 or None, a random seed will be used.
  • topK (integer): Limits sampling to the top K most probable tokens; defaults to 250.
  • topP (number): Controls sampling to tokens with a cumulative probability of P; defaults to 0.
  • prompt (string): A description of the musical theme or style you wish to generate (e.g., "a piano solo").
  • duration (integer): The length of the generated audio in seconds; default is 8 seconds.
  • sourceAudio (string): A URI linking to an audio file that influences the generated music.
  • temperature (number): Adjusts the diversity of sampling; higher values yield more varied outputs.
  • continuation (boolean): If true, the generation will continue from the provided source audio.
  • audioOutputFormat (string): Specifies the format for the output audio; options are 'wav' or 'mp3' (default is 'wav').
  • continuationEndTime (integer): Specifies the end time for audio continuation.
  • inputAdherenceFactor (integer): Influences how closely the output matches the input; higher values lead to closer adherence.
  • continuationStartTime (integer): Specifies where to start the continuation from the source audio.
  • useMultiBandDiffusion (boolean): If true, uses MultiBand Diffusion for decoding (non-stereo models only).
  • generationModelVersion (string): Selects the version of the model for music generation; options include "stereo-melody-large", "stereo-large", "melody-large", "large".
  • audioNormalizationStrategy (string): Defines how to normalize the generated audio; defaults to 'loudness'.

Example Input:

{
  "prompt": "a piano solo ",
  "duration": 60,
  "generationModelVersion": "large"
}

Output

The output from the action will typically return a URI link to the generated audio file, which you can play or download. An example response may look like this:

Example Output:

"https://assets.cognitiveactions.com/invocations/da0f9e47-8a95-40bf-80c8-bbdfe7772add/f0e86ac2-101b-468c-9023-396782530a65.wav"

Conceptual Usage Example (Python):

Below is a conceptual Python code snippet demonstrating how to call the MusicGen action. Replace the placeholder values with your actual API key and action ID.

import requests
import json

# Replace with your Cognitive Actions API key and endpoint
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute" # Hypothetical endpoint

action_id = "f8e54cd6-38ea-4763-b1aa-b5a44b3dca3e"  # Action ID for Generate Music with MusicGen

# Construct the input payload based on the action's requirements
payload = {
    "prompt": "a piano solo ",
    "duration": 60,
    "generationModelVersion": "large"
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json"
}

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json={"action_id": action_id, "inputs": payload} # Hypothetical structure
    )
    response.raise_for_status()  # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body: {e.response.text}")

In this code example:

  • Replace the COGNITIVE_ACTIONS_API_KEY with your actual API key.
  • The action_id corresponds to the Generate Music with MusicGen action.
  • The payload is structured based on the required input schema, ensuring that you specify the necessary parameters for the generation.

Conclusion

The Charles McCarthy MusicGen Cognitive Action presents an exciting opportunity for developers to generate unique audio compositions with ease. By leveraging the customizable parameters available, you can create music that fits your specific needs, from classical piano solos to experimental soundscapes. As you integrate these actions into your projects, consider exploring various combinations of parameters to discover the full potential of automated music generation. Happy coding, and may your applications resonate with creativity!

← Back to All Articles