Arc Search · early access
Building the AI-native workflow for audio professionals
An AI-native workflow that keeps control in the hands of creators, while accelerating the non-creative work. Our first product, Arc Search, finds the right files in your library from a plain-language description or a reference clip. All entirely local to your device. Your files are never uploaded, modified, or trained on.
- glass shattering on concrete
- distant thunder clap
- heavy creaking door
- large creature guttural snarl
- footsteps on wet gravel
- metallic impact with a long tail
- rain drumming on a tin roof
- car engine sputter
- Similar to: MACK Whoosh Synth Rumbling.wav
- glass shattering on concrete
- distant thunder clap
- heavy creaking door
- large creature guttural snarl
- footsteps on wet gravel
- metallic impact with a long tail
- rain drumming on a tin roof
- car engine sputter
- Similar to: MACK Whoosh Synth Rumbling.wav
Arcantus
AI handles the groundwork You keep creative control
Arcantus builds AI-native tools for professional sound design. We are reimagining how creators work with AI technology by embedding it into workflows that emphasize taste and creativity. Arc Search is built from the bottom up for the slowest part of sound design: finding the right source material in your own library. Our goal is to make the most efficient sound search tool available. So you can find better sounds faster and spend more time creating.
Arc Search - now in early access
0:10-0:19
The exact file and take you need
Every result opens on an editable waveform. Pull one take from a long recording, shift pitch and playback rate independently, loop the region, and zoom in close. What you audition is what exports.
47,745 files · 192kHz / 2ch
No limits on what you find
Arc Search returns every file that matches, with no cap on results and no restrictions on file type, sample rate, channels, or length. It is built for your library and the way you work as a sound designer.
100% local
Nothing leaves your machine
Indexing and search run on your device and work offline. Your files are never uploaded, modified, or trained on.
Describe it, or drop a clip
One slider covers multiple ways to find a sound. One end runs traditional keyword and metadata search while the sonic end takes a plain-language description to match on context. Set it anywhere in between to blend the two. Have a reference sound or an example result? Drop it into the search bar to find results that match how it sounds. Whichever way you search, every result shows how it matched.
Try it on your own library.
Arc Search is in early access. Request access and put it to work on your own recordings.