suno-music_skill

Overview

This skill generates full AI songs with vocals using Suno's Sonic engine through the MusicAPI.ai wrapper. It produces MP3/MP4/WAV outputs, supports Sonic V3.5–V5 models, and returns two variants per generation. An API key from MusicAPI.ai is required and each song call consumes credits.

How this skill works

The skill makes REST calls to MusicAPI.ai to create a Sonic generation task, polls the task until completion, and returns direct download URLs for the final audio and optional video. It supports custom-mode (you supply structured lyrics) or AI-description mode (AI generates lyrics and arrangement), plus options for instrumentals, vocal gender control, and stem separation. Node.js examples and polling logic are provided for automation.

When to use it

• You want a full song with AI vocals quickly and without managing a Suno account.
• You need both instrumentals and vocal versions or separate stems for mixing.
• You want to prototype song ideas from a short description or full lyrics.
• You plan to submit generated tracks to a music platform and need ready-to-upload MP3s.
• You need explicit control over model version or vocal gender.

Best practices

• Use sonic-v5 for highest quality; fallback to older models only if needed.
• Provide clear, comma-separated tags for genre, mood, vocals, and instruments to steer style.
• Use custom_mode:true with structured lyrics ([Verse], [Chorus], etc.) for tight control.
• Poll task status every 15–25 seconds and handle failed/timeout states gracefully.
• Store the MusicAPI key in an environment variable and never hardcode it.

Example use cases

• Generate a pop single from a short concept using sonic-v5 and submit the MP3 to a distribution platform.
• Create a lo-fi instrumental for background use by setting make_instrumental:true.
• Produce two vocal variants of a song, compare mixes, and request stem separation for remixing.
• Extend an existing song by sending an extend_task_id to continue the arrangement.
• Automate batch generation from descriptions for rapid prototyping of song catalogs.