Instrumental Mode generates music without vocals: a full arrangement with instrumentation, but no sung or spoken vocal line. This guide covers what the mode does, how to run it, and where the limits are.
- Turns vocals off for the task. The workbench sends an instrumental arrangement request instead of a vocal song.
- Keeps everything else. Style, tempo, mood, instrumentation and structure are still driven by your prompt and the fields you fill in Custom Mode.
- Still costs credits. Instrumental tasks are billed like any other task, and the cost of the selected model is shown before you submit.
- Open the AI Music Generator and sign in — authentication is required before a task can be submitted.
- Switch to Custom Mode so you can control Title, Style of Music, Lyrics, the Instrumental toggle and the model in one form.
- Enable Instrumental. Leave Lyrics empty — the form expects lyrics only when instrumental mode is off.
- Describe the arrangement in Style of Music: genre, era, tempo, and the instruments you want, for example
cinematic piano, warm strings, slow build, 90 bpm.
- Pick the model and read the credit cost shown next to it.
- Submit and review. Play the result in the workbench, then download the finished track from the result panel.
- Background beds for videos, podcasts, tutorials or streams where a vocal would compete with speech.
- Practice and rehearsal material: a backing arrangement you can play over.
- Placeholder music during editing, before you commit to a licensed track.
- Anyone who wants the arrangement first and will add vocals later with the Add Vocals tool.
- No MIDI and no per-instrument stems by default. You get one mixed track. The workbench does offer vocal separation (two stems or multiple stems) as a separate task, but that is separation, not transcription into notes or a DAW project.
- No guarantee that a specific instrument is isolated. Ask for a sparse arrangement if you need one element to stand out.
- No automatic commercial rights. Commercial rights are not automatically granted; eligibility depends on the selected model provider, your plan, your inputs, applicable law and third-party rights. Check the Terms of Service.
- No fixed number of generations. Each task consumes credits, and heavier models cost more.
- Leaving Style of Music empty. The instrumental toggle removes vocals; it does not choose a genre for you.
- Describing a vocalist, then wondering why there is no voice. In instrumental mode, vocal directions are ignored at best.
- Rewriting the whole prompt after a near miss. Change one variable at a time — tempo, instrumentation, or dynamics.
- Expecting loudness or length to be exact. Length and mix depend on the model; extend the track if you need it longer.
Yes — Instrumental Mode is the switch for generating music without vocals. If you later want a vocal version, the Add Vocals workflow adds a vocal layer to an existing track.
Yes, as a separate task: the Vocal Remover separates a generated track into vocal and instrumental parts. It is a different tool from instrumental generation, so the result is split audio rather than a newly composed arrangement.
Because every generation is a billed task on the platform: the workbench needs an account to attach credits, show your balance and store the result.
Use Extend in the workbench: pick the source track, set the point to continue from in seconds, and describe the next section. Extending is another task, so it consumes additional credits.
Open the AI Music Generator →
