Create My Song
Tell AI what you feel. Hear a song built around your voice.
You don't need lyrics.
You don't need composition.
Tell the AI what you're thinking. It can interview you, recommend a direction, write an original song, fit the composition to your voice and prepare the generation pipeline.
What are you feeling or thinking?
Speak naturally or type. A few rough thoughts are enough.
Teach the studio your voice.
Record a comfortable singing sample. The app stores only what you choose to upload.
Voice sample
Sing 30–60 seconds comfortably in your natural voice. A clean solo vocal sample gives the best voice-reference result.
Projects
Your song workspace. Main Admin can see all users' projects.
User Rights
Create users, activate/deactivate access and control exactly what each person can use.
Add user
Setup
Free-first architecture with Supabase persistence, private audio storage, PWA install and a real local ACE-Step vocal worker.
1. App Server
npm startNode server on Render Free.
2. Supabase
DATA_BACKEND=supabaseUsers, rights, sessions, projects, voice profiles and jobs persist in Supabase.
3. Private Audio
music-uploads + music-generatedVoice samples and generated tracks stay in private Supabase Storage buckets.
4. Real AI Music Worker
AI_MODE=workerYour Windows/GPU computer runs ACE-Step locally, generates full songs with vocals and uploads them back to this app.