v1.3.0 — Subtitle-style transcript + faster, safer export

What's new

The transcript editor now reads as short subtitle-style lines with timestamps instead of a wall of paragraph text. Much easier to scan, navigate, and delete from. And the export sheet finally lets you cancel, with an honest progress indicator that doesn't lie at 0%.

Changes

• Subtitle-style transcript — each line is a sentence, up to 10 words, or a silence break. Small timestamp per line for orientation. Editing still works the same (click, drag, shift-click, cmd-click, Delete).
• Fixed audio/video scatter on deletions — deleting runs of words now cuts everything between them, including the tiny inter-word gaps Whisper's timestamps leave behind
• Cancel Export button — kills the FFmpeg subprocess instantly instead of forcing a quit
• Honest progress bar — shows activity while FFmpeg sets up, flips to real percent once encoding begins
• Smoother cuts by default — audio crossfade bumped from 70 ms to 200 ms
• Under the hood — ~500 lines deleted, modernized to native AppKit patterns (dynamic NSColors, native event handling, AVPlayer periodic observer), removed deprecated APIs

Install

Download Redact.dmg, open it, drag Redact to Applications.
Requires macOS 14+ · Apple Silicon & Intel

v1.2.0 — Native WhisperKit Transcription

What's new

Transcription now runs entirely on-device via WhisperKit (CoreML + Metal). No more Python, pip, or virtual environments — the app is fully self-contained. Just install FFmpeg and go.

Changes

• Native transcription — replaced Python/faster-whisper subprocess with WhisperKit. Models download automatically on first use
• More model options — added English-only variants (Tiny, Base, Small, Medium) and Large v3 Turbo
• Simpler setup — the only prerequisite is now brew install ffmpeg
• Smaller CoreML models — optimized model sizes (e.g., Small is ~250 MB vs ~465 MB before)

Install

Download Redact.dmg, open it, drag Redact to Applications.
Requires macOS 14+ · Apple Silicon & Intel

Redact 1.1.0

What's new

• Card-based Preferences — Settings reorganized into rounded-rect cards (Appearance, Transcript, Audio, Transcription, ElevenLabs Voice)
• Theme support — Dark, Light, and System theme with full color system
• Customizable transcript — Choose font family, size, and highlight color
• Crossfade control — Adjustable audio fade duration at cut boundaries
• Whisper model management — Select and download models directly from Preferences
• ElevenLabs voice integration — API key, voice selection, and custom voice ID for TTS export
• Export improvements — MKV/WebM support, quality and speed presets, ElevenLabs voice recreation
• Refreshed UI — Themed surfaces, dividers, and text hierarchy across all views

Install

1. Download Redact-1.1.0.dmg below
2. Open the DMG and drag Redact to Applications
3. Launch Redact

Signed and notarized by Apple.

Requirements

• macOS 14+
• FFmpeg (brew install ffmpeg)
• Python 3 + faster-whisper (pip3 install faster-whisper)

Redact 1.0.0

Redact 1.0.0 — Initial Release

Native macOS rewrite of Redact. Remove unwanted words from video by clicking them in the transcript.

What's included

• Import video/audio files (MP4, MKV, WebM, MOV, AVI, MP3, WAV, M4A)
• Automatic transcription via faster-whisper with word-level timestamps
• Click, drag, shift-click, cmd-click word selection
• Real-time preview with 70ms audio crossfades at cut points
• Waveform visualization with click-to-seek
• Unlimited undo/redo
• Export to MP4/MKV/WebM with quality and speed options
• Export SRT subtitles (timestamps adjusted for deletions)
• Save/load .rdt project files

Requirements

• macOS 14 (Sonoma) or later
• Apple Silicon or Intel
• FFmpeg (brew install ffmpeg)
• Python 3 + faster-whisper (pip3 install faster-whisper)

Install

1. Download Redact-1.0.0.dmg below
2. Open the DMG and drag Redact to Applications
3. Since this build is not notarized, right-click the app > Open on first launch

Size

~3.5 MB DMG (vs ~200 MB Electron version)