Skip to content

FlowScribe v0.2.3

Choose a tag to compare

@github-actions github-actions released this 14 May 04:12
· 169 commits to main since this release

FlowScribe v0.2.3

Windows portable release for FlowScribe.

What's Included

  • Local audio/video transcription CLI.
  • PySide6 desktop GUI for transcription, transcript viewing, search, and media sync.
  • Public URL transcription with audio-first download/extraction.
  • Local and URL media inspection with flowscribe inspect.
  • Keyword search and timestamp location with flowscribe search.
  • TXT, Markdown, JSON, SRT, and VTT transcript output.
  • Chinese, English, and mixed-language workflows.
  • Word timestamp data for future text-to-video navigation.
  • Explicit cookies and proxy support for URL media access.
  • flowscribe doctor environment diagnostics.
  • Chinese-oriented preset: --preset zh.
  • Portable Windows one-folder build.
  • Bundled ffmpeg.exe and ffprobe.exe.

Download

Download one of these Windows x64 portable packages:

  • FlowScribe-v0.2.3-windows-x64.zip for the CLI build
  • FlowScribeGUI-v0.2.3-windows-x64.zip for the desktop GUI build

CLI quick check:

.\FlowScribe\FlowScribe.exe doctor

GUI quick start:

.\FlowScribeGUI\FlowScribeGUI.exe

Notes

  • Windows x64 portable build.
  • Whisper models are not bundled in the executable.
  • The first transcription with a selected model may download model files from Hugging Face.
  • The GUI release defaults to quiet user logging mode and launches without a console window.
  • tiny is recommended only for smoke tests.
  • small is the recommended starting model for real use.
  • Use medium or larger models when accuracy matters more than speed.

Legal Boundary

FlowScribe is intended for personal learning, accessibility, research notes, and lawful information processing. Do not use it to bypass DRM, crack applications, or redistribute copyrighted transcripts without permission.