Skip to content

Utterleaf v0.3.6

Choose a tag to compare

@github-actions github-actions released this 09 Sep 17:01
· 256 commits to main since this release

A quieter, friendlier Utterleaf. v0.3.6 refreshes Settings and the documentation without adding a UI framework or changing the speech engine.

  • Charcoal surfaces, grouped controls, softer field borders, and the approved wordmark in the sidebar.
  • Clearer headings and Help organized around recovery, device checks, and optional artwork.
  • A simpler GitHub download table, theme-aware branding, a documentation index, and refreshed real-app screenshots.
  • Historical reviews are clearly marked. The existing single Settings window, staged defaults reset, and Wayland recovery fixes are retained.

Download and open

Computer Archive After extracting the whole archive
Windows x64 Utterleaf-windows-x64-cpu.zip Open Utterleaf/utterleaf.exe
macOS Apple Silicon Utterleaf-macos-arm64.tar.gz Run ./Utterleaf/utterleaf
Linux x64 (Ubuntu 24.04 or compatible) Utterleaf-linux-x64.tar.gz Run ./Utterleaf/utterleaf

Python is bundled. Keep _internal beside the executable. The selected speech
model downloads on first launch; recognition then runs locally. Model weights
and optional CUDA libraries are separate. SHA256SUMS.txt lists archive checksums.

Platform setup

  • Windows: Ctrl+Win is the default shortcut. GPU acceleration needs compatible CUDA libraries; the CPU build works without them.
  • macOS: Apple Silicon only, with CPU inference. Grant Microphone and Accessibility permissions. The portable build is unsigned and not notarized.
  • Linux: X11 needs a paste helper. Wayland requires XWayland, a desktop shortcut to the absolute executable path plus --toggle, wl-clipboard, and a compatible/configured paste helper. Native Wayland dictation still needs desktop-specific validation.

These are portable archives, not installers. All builds are unsigned. Source tests
and packaged smoke checks do not establish native microphone and editor delivery
on every desktop. Long dictation, meetings, system-audio captions, and translation
remain planned features.

Installation ·
Wayland help ·
Screenshots ·
Earlier releases