Skip to content

Releases: seeknull/openwispr

Release list

v0.1.1

v0.1.1 Latest
Latest

Choose a tag to compare

View all tags
@gurugv gurugv released this 14 May 15:18
v0.1.1
e4bdf6c

OpenWispr v0.1.1

First point release after v0.1.0.

What's new

  • Live mic-level visualizer in the HUD. The listening pill now shows a pulsing red dot plus 3 small bars that move with your voice — you can see at a glance that the mic is hearing you, not just that the app says it is.
  • Settings shows the running build. Footer line on every Settings tab (and a fuller box on About) shows version, build timestamp, and git SHA. Useful for bug reports.
  • Fixed: "Keystroke Receiving" prompt loop. Posting events without Accessibility used to trigger macOS's prompt repeatedly per transcribed line. TextInjector now pre-checks AXIsProcessTrusted() and silently no-ops if the grant is missing.
  • HUD polish. Smaller, more translucent, properly self-sizing so the label can never clip. Bottom-center placement out of the way of menu bars and editor chrome.

Fixed

  • HUD crash on the second toggle (SwiftUI animation outliving the host window).
  • HUD layers rendering under the blur view (dot and bars invisible).
  • AudioLevelMeter crashing because the audio thread touched a @MainActor class.

Install

  1. Download OpenWispr.zip below.
  2. Unzip and drag OpenWispr.app to /Applications.
  3. Right-click → Open the first time (the build is unsigned; one-time Gatekeeper acceptance).
  4. Click the menu-bar waveform icon → Settings → grant Microphone and Accessibility.
  5. Press Fn + Option anywhere — speak — press again to stop.

If you're upgrading from v0.1.0, replace OpenWispr.app in /Applications with this build. TCC permissions may need re-granting since the code signature changes per build — Settings → Permissions has a Hard Reset button if it gets stuck.

Privacy

OpenWispr is fully on-device — no network calls in the dictation path. Verify yourself:

grep -r "URLSession\|http://\|https://" Sources/

Requirements

  • macOS 13 (Ventura) or newer
  • Apple Silicon Mac (arm64)

Checksums

See OpenWispr.zip.sha256 below.

Assets 4
Loading

v0.1.0

v0.1.0

Choose a tag to compare

View all tags
@gurugv gurugv released this 14 May 14:34
v0.1.0
9a9e67a

OpenWispr v0.1.0 — initial release

Speak. The text appears where your cursor is.
Free, open source, fully on-device dictation for macOS.

Install

  1. Download OpenWispr.zip below.
  2. Unzip and drag OpenWispr.app to /Applications.
  3. First launch: right-click OpenWispr.app → Open (the build is unsigned; this is a one-time Gatekeeper acceptance).
  4. Click the menu-bar waveform icon → Settings → grant Microphone and Accessibility.
  5. Press Fn + Option anywhere — speak — press again to stop. Text appears at the cursor.

Full walkthrough: docs/install.md.

Privacy

OpenWispr is fully on-device — no network calls in the dictation path. Verify yourself:

grep -r "URLSession\|http://\|https://" Sources/

Only matches are a link to this repo in the About window.

Requirements

  • macOS 13 (Ventura) or newer
  • Apple Silicon Mac (arm64). Intel builds will land in a future release.

What's included

  • Pre-bundled Moonshine medium-streaming-en model (~280 MB) — no separate download needed.
  • Two text-insertion modes (clipboard paste, synthesized keystrokes).
  • Self-test that surfaces permission / hotkey problems via a colored menu-bar icon.

Known limitations

  • Hotkey is hard-coded to Fn+Option (rebinding lands in v0.2).
  • Ad-hoc signed: TCC permission grants don't survive across reinstalls (re-grant via Settings → Permissions if you hit this).

Checksums

See OpenWispr.zip.sha256 below.

Loading