GLOAM.FM
VOICE STUDIO
Features Models Intelligence Voice Lab Developers Docs Changelog Pricing Download
CHANGELOG

What's new

Every release of Gloam Voice Studio, newest first. Full notes live on GitHub Releases.

v0.1.0-build.4 2026-07-03 PREVIEW

Gloam Voice Studio — a local, on-device text-to-speech studio for macOS (Chatterbox, Qwen3-TTS, Fish). This build ships the Chatterbox "Regular" quality fix, bringing the Swift/MLX port to PyTorch-reference parity (resampler anti-aliasing + fresh-noise fixes).

Download

GloamVoiceStudio-0.1.0-macOS.zip — unzip to get GloamVoiceStudio.app.

Requirements
  • macOS 14.0 (Sonoma) or later
  • Apple Silicon (M1 or later) recommended — universal binary, but MLX inference is built for Apple Silicon
  • Internet on first use: TTS models download on demand from Hugging Face
First launch (important)

This is an ad-hoc-signed build (not Apple-notarized — there's no Developer ID cert on hand), so Gatekeeper blocks the first open. One-time step:

1. Unzip and move GloamVoiceStudio.app to /Applications
2. Right-click the app → Open → Open in the dialog

If macOS still refuses ("damaged / cannot be opened"), clear the quarantine flag in Terminal:

xattr -dr com.apple.quarantine /Applications/GloamVoiceStudio.app

After the first open it launches normally.

Notes
  • The App Store build (identical code) is in review; once approved it installs with no Gatekeeper step.
  • This is a preview build for testing — please report issues.
Follow development
Watch the repo for release notes as they land.
View releases on GitHub