Skip to content
Inworld AI Blog· The Inworld team·· 4 d agoPickAI score55

Inworld open-sources voice isolation model Voice Isolator with offline and real-time versions

Introducing Voice Isolator

AI summary

Inworld released Voice Isolator, its first open-source model, which removes background voices, music, and transient noise while preserving the speaker's timbre, breathing, and volume. It works at 48 kHz and comes in two checkpoints: offline and real-time streaming with 20 ms lookahead.

Why it matters

Inworld open-sourced its first voice isolation model, offering 48 kHz dual versions and a side-by-side comparison on its self-built VIBE benchmark, which can help assess whether it fits voice agents or recording processing.

Source: Inworld AI Blog · inworld.ai