Generate sounds from words. Directly in your DAW.
Featured in Ableton's blog post, AI and Music-Making.
Checkout our latest project, Vroom Live.
In the last few months the field of text-to-audio AI models has rapidly evolved. VroomAI is a VST plugin delivering the latest text-to-audio AI models directly to artists.
VroomAI was created as a submission to the 2023 Neural Audio Plugin Competition.
Audio samples are saved into a user-specified directory and can be played at various pitches inside the DAW.
- AudioLDM (Haohe Liu, Zehua Chen, Yi Yuan, Xinhao Mei, Xubo Liu, Danilo Mandic, Wenwu Wang, Mark D. Plumbley).
- MusicLM (Internal implementation coming soon)
At the moment the quickest way to get things working is to directly build the VST on your machine so we don't need to worry about code signing. A public inference server will be available shortly at vroomai.com but in the meantime you can host your own using Haohe Liu's app.py with the api enabled.
This plugin has only ever been tested on Reaper and Ableton on MacOS (M2).
If you want to ask any questions or follow updates - join the discord!
Barney Hill (University of Oxford)