Skip to content

Releases: wolfyy970/audio-monster

Audio Monster v0.2.3

Choose a tag to compare

@wolfyy970 wolfyy970 released this 21 Jul 05:12

Full Changelog: v0.2.2...v0.2.3

Audio Monster v0.2.2

Choose a tag to compare

@wolfyy970 wolfyy970 released this 20 Jul 05:51

Full Changelog: v0.2.1...v0.2.2

Audio Monster v0.2.1

Audio Monster v0.2.1 Pre-release
Pre-release

Choose a tag to compare

@wolfyy970 wolfyy970 released this 20 Jul 01:43

Audio Monster 0.2.1

This release updates the native article-extraction pipeline to SwiftReadability 0.3.1 while keeping publisher-specific and media-recovery behavior explicitly owned by Audio Monster. Existing extraction behavior remains intact, with no Python or JavaScript runtime required.

The download is a signed and Apple-notarized build for Apple Silicon Macs running macOS 14 or later. Verify Audio-Monster-v0.2.1-macOS-arm64.zip with the accompanying SHA-256 file before opening it.

Full changelog: v0.2.0...v0.2.1

Audio Monster v0.2.0

Audio Monster v0.2.0 Pre-release
Pre-release

Choose a tag to compare

@wolfyy970 wolfyy970 released this 19 Jul 22:40

What's Changed

  • Release v0.2.0 with native Swift Readability by @wolfyy970 in #3

New Contributors

Full Changelog: v0.1.0...v0.2.0

Audio Monster v0.1.0

Audio Monster v0.1.0 Pre-release
Pre-release

Choose a tag to compare

@wolfyy970 wolfyy970 released this 19 Jul 10:43

Audio Monster 0.1.0 is the first public preview of the native macOS menu-bar app that turns readable web articles into locally narrated audio.

Install

  1. Download Audio-Monster-v0.1.0-macOS-arm64.zip below.
  2. Optionally verify it with the accompanying .sha256 file.
  3. Unzip it, move Audio Monster.app to Applications, and launch it normally.

The app is signed with Developer ID and notarized by Apple. This build requires macOS 14 or later on an Apple Silicon Mac. Intel Macs are not supported because MLX and MLX Audio Swift target Apple Silicon.

Highlights

  • Native Swift Kokoro speech synthesis through MLX and Metal; no Python service.
  • Compressed M4A/AAC output with the source URL embedded in metadata.
  • Fifty-four grouped voices with automatically generated previews and selection autoplay.
  • Progressive playback and pitch-preserving speed control from 0.2× to 3×.
  • Mozilla Readability extraction inside WebKit.
  • iCloud Documents storage with local and custom-folder options.
  • Saved-file library with playback and Finder integration.

The first synthesis downloads the Kokoro model from Hugging Face. Text extraction and speech inference otherwise remain on the Mac.

SHA-256: 8838446194a06501d0ce206bc977b18d16442b53421c409922eb54072cb5170f

See the changelog and full commit history.