Skip to content

Latest commit

 

History

History
105 lines (83 loc) · 3.29 KB

README.md

File metadata and controls

105 lines (83 loc) · 3.29 KB
Raw

42Summaries

42Summaries is a macOS application that transcribes audio and video files, then generates concise summaries using various LLM providers including Ollama, OpenAI, and Anthropic.

Testflight

Join Beta

AppStore

Download on the App Store

Features

Core Features

  • Audio and video file transcription using WhisperKit
  • Multiple LLM providers for summarization:
    • Ollama (local)
    • OpenAI
    • Anthropic Claude
  • Power mode settings for transcription (fast vs. energy efficient)
  • Videos will be converted to audio files for transcription
  • Export capabilities:
    • PDF export
    • TXT export
    • Microsoft Teams integration
    • Direct export to ChatGPT, Claude, and Perplexity

UI Features

  • Modern SwiftUI interface with sidebar navigation
  • Dark mode support
  • Progress tracking for transcription
  • Real-time notifications
  • Customizable prompt library
  • Format controls for summaries
  • Error handling and user feedback

Project Structure

Core Components

  • Services
    • TranscriptionService: Handles audio/video transcription using WhisperKit
    • SummaryService: Manages LLM integrations (Ollama, OpenAI, Anthropic)
    • ExportManager: Handles various export formats and destinations
    • TeamsAuthManager: Manages Microsoft Teams integration
    • NotificationManager: Handles system notifications

Views

  • Main Views
    • WelcomeView: Initial user onboarding
    • FileSelectionView: Audio/video file selection
    • TranscriptionView: Transcription process and results
    • SummaryView: Summary generation and editing
    • SettingsView: App configuration and preferences
  • Supporting Views
    • LaunchScreenView: App initialization
    • AboutView: App information
    • ExportOptionsView: Export functionality
    • Error/NotificationView: User feedback
    • TeamsSelectionView: Teams integration interface

Requirements

  • macOS 15.0+
  • Xcode 16.0+
  • Swift 6.0+

Installation

  1. Clone the repository
  2. Open the project in Xcode
  3. Build and run the application

Dependencies

  • SwiftUI for the user interface
  • WhisperKit for transcriptions
  • OllamaKit for local LLM integration
  • LLMChatOpenAI for OpenAI integration
  • LLMChatAnthropic for Claude integration
  • MSAL for Microsoft Teams authentication

Configuration

LLM Provider Setup

  • Ollama: Requires local Ollama installation
  • OpenAI: Requires API key in settings
  • Anthropic: Requires API key in settings

Microsoft Teams Integration

  • Requires Azure AD application registration
  • Configure Client ID and Tenant ID in settings

Getting Started

  1. Launch the application
  2. Follow the getting started guide at https://42summaries.com/#get-started
  3. Configure your preferred LLM provider in Settings
  4. Select an audio/video file to begin transcription
  5. Generate and customize summaries as needed

Additional Notes

  • The app uses WhisperKit for offline transcription
  • Supports multiple export formats and destinations
  • Includes a customizable prompt library for different summary styles
  • Features both fast and energy-efficient transcription modes