Hush is a macOS app that uses on-device AI to remove noise and reverb from spoken audio, trusted by podcasters, audiobook narrators and voice actors who record in less-than-perfect rooms.
Most noise reduction tools rely on EQ and compression tricks that leave voices sounding thin or watery. Hush instead uses a transformer-based neural network trained to separate speech from broadband noise, transient sounds like clicks and rustling, and the reverberation of an untreated room. It runs entirely on your Mac’s Apple Neural Engine, so processing is fast, your recordings are never uploaded anywhere, and it works fully offline. Batch processing handles long-form narration and podcast back-catalogs, and the companion Hush Pro plug-in brings the same engine into Pro Tools for professional post-production workflows.
A free 21-day trial includes every feature. The Hush app itself is a one-time purchase of $99, while the Hush Pro plug-in for Pro Tools costs $249 one-time. There are no subscriptions and no per-minute fees.
Hush was created by developer Ian Sampson out of frustration with what traditional noise reduction did to his own recordings, and it has earned a strong word-of-mouth reputation among professional narrators. Because there is no account, no subscription and no cloud processing, it fits client work where recordings cannot leave the machine ā and replaces what previously meant renting treated studio space.
Podcasters, audiobook narrators, voice actors and video creators on Mac who record at home and need studio-clean speech without building a treated studio.