Describe your scene. Get the sound.
Type what is happening at your table and we’ll build a playable mix from 475 music tracks, 283 ambience beds and 533 sound effects. Free with an account — and if it isn’t right, change any layer yourself.
Why not just pick a scene?
Because the scene you need is usually not on anyone’s list of 100. A courtroom where a jury decides your fate, a hospice on the edge of a plague quarter, a negotiation in a burning library — those are the moments a session turns on, and they are exactly the ones no pre-made soundboard covers.
The trick is that the model never picks the audio. It reads your description and produces search terms; a deterministic matcher then finds real files in the library. So you always get sounds that exist, never a plausible-looking list of things we don’t have.
AI scene builder — FAQ
Is it free?
Yes, with a free account — each build runs a real request against the library, so we ask you to sign in first. Everything it produces is yours to play, edit and keep.
Does it generate the audio?
No. Everything you hear is from our own library of composed music and recorded effects. The AI only decides which of them fit your description, so nothing is synthesised and nothing is invented.
What if the mix is wrong?
Change it. Every layer can be swapped, muted or re-levelled, and you can upload your own audio too — all free.
Prefer to start from something ready-made? Browse all 100 scenes →