Start with what you have

Use the built-in synth instrumentals, your phone or computer microphone, and headphones. Recording, effects, saving and WAV exports need no API account.

• Choose a genre, key and tempo that fit your voice.
• Record in a quiet, soft-furnished room.
• Keep the backing quieter than your singing.
• Save a project backup after each session.

Spend on the missing piece

OpenAI can help with original lyrics, song structure, instrumental prompts and transcribing a take. It is not the music generator here. Singing transcription may need corrections.

ElevenLabs creates AI instrumentals. Start with a 30-second test before requesting a full track. You can also import an instrumental from any service you have permission to use.

API billing is separate from ChatGPT subscriptions. Costs, music access and release rights depend on your provider and plan.

A better take beats more plugins

Wear wired headphones where possible. Bluetooth adds delay; use the vocal offset if your take feels late. Stand a consistent distance from the mic and avoid clipping.

Check the microphone meter, sing the loudest line, then adjust your microphone's input level in device settings. Aim for healthy green levels rather than red.

The app offers low-cut filtering, compression, EQ and reverb. Advanced pitch correction, stem separation and professional mastering need separate tools.

Keep the parts and the project

Download the full WAV mix, instrumental and dry voice separately. Keep your dry vocal so you can remix it later. Project backups include audio, lyrics, settings and takes.

Browser storage is local, can be cleared, and is not a cloud backup. Projects don't sync between devices automatically.

Recording requires HTTPS or localhost and microphone permission. Modern Chrome, Edge, Safari and Firefox are the intended browsers. Mobile performance depends on available memory.