Vocal Gate Free Download: AI-Powered Speech Detection Gate by Dan K
Dan K’s Vocal Gate is a free, open-source VST3/AU noise gate plugin that uses an AI model to detect speech and automatically duck unwanted microphone artifacts — including coughs, sneezes, keyboard noise, and heavy breathing. It is designed for podcasters, streamers, and video editors, as well as studio use in a DAW.
Key Features
- AI model for speech detection: automatically ducks non-speech artifacts (coughs, sneezes, keyboard noise, breathing)
- Supported OS: macOS 11 and later (Apple Silicon), Windows 10 and later (x64), Linux (Ubuntu 22.04 and later)
- Supported formats: VST3 (all OS), AU (macOS only)
- Verified compatible with OBS Studio, Streamlabs, vMix (streaming), DaVinci Resolve, Adobe Premiere Pro, Final Cut Pro (video editing), Reaper, Logic Pro, FL Studio, Ableton Live, Audacity (DAW)
- Supports offline rendering (bouncing)
- AI model uses an int8 quantized ONNX model for approximately 0.3ms real-time inference per buffer
- Open source on GitHub
Important Note for Live Streaming
Vocal Gate requires a 750ms lookahead for the AI model to make predictions. When using it in OBS Studio or other live streaming software, you must delay your video source and desktop audio by 750ms to keep audio and video in sync. For offline use in a DAW or video editor, no additional setup is required.
How to Download Vocal Gate for Free
Download the installer for your platform from the GitHub Releases page. Detailed setup instructions for OBS and other applications are available in the README on the GitHub page.