ScreenRecAI-Desktop – Screen Recorder with Local AI Transcripts & Summaries

An open-source desktop screen recorder built with Electron, React, and TypeScript. Capture your screen, save videos locally, and auto-generate transcripts and AI summaries using Ollama — all offline, privacy-first, and fully customizable.

✨ Features

• 🎥 Screen Recording - Capture your entire screen or specific windows
• 📁 Smart File Organization - Automatically saves to organized folders with timestamps
• 🎵 Audio Extraction - Extracts audio from recorded videos for processing
• 📝 AI Transcript Generation - Converts speech to text using local AI models
• 🤖 Intelligent Summarization - Generates context-aware summaries using Ollama
• 🎯 Multiple Recording Types - Optimized for Google Meet, Lessons, and general videos
• 👀 Live Preview - See what you're recording in real-time
• 🔄 Cross-Platform - Works seamlessly on Windows, macOS, and Linux
• 🔒 Privacy-First - All processing happens locally, no data sent to external servers
• ⚡ Automated Setup - One-command setup for all binary dependencies

📸 Screenshots

Main Application Interface

Recording in Progress

🚀 Quick Start

Download Pre-built Binaries

The fastest way to get started:

1. Visit Releases
2. Download for your platform:
   • Windows: meetingvideo-transrecorder-1.0.0-setup.exe
   • macOS: meetingvideo-transrecorder-1.0.0.dmg
3. Install and launch the application

Build from Source

For developers or those who need the latest features:

# Clone the repository
git clone [https://github.com/LinuxDevil/ScreenRecAI-Desktop](https://github.com/LinuxDevil/ScreenRecAI-Desktop)
cd meetingvideo-transrecorder
# Install dependencies
npm install
pip install -r requirements.txt
# Start development server
npm run dev

📋 Requirements

Essential Dependencies

Component	Version	Purpose
Node.js	16+	Runtime environment
Python	3.8+	Audio processing
FFmpeg	Latest	Video/audio conversion

Optional (for AI features)

Component	Purpose
Ollama	Local AI model for summarization
Python packages	Speech-to-text processing

🛠️ Setup Guide

Automated Setup (Recommended)

The easiest way to set up all binary dependencies:

# Setup for current platform (recommended)
./setup-binaries.sh
# Or setup for all platforms
./setup-binaries.sh all
# Or setup for specific platform
./setup-binaries.sh macos 
./setup-binaries.sh windows
./setup-binaries.sh linux

What the Setup Script Does

macOS Setup

• Creates a Python virtual environment in python-runtime/
• Installs all dependencies from requirements.txt
• Downloads FFmpeg binary to ffmpeg-bin/

Windows Setup

• Downloads FFmpeg binary to ffmpeg-bin-windows/
• Uses system Python (no bundled runtime needed)
• Requires pip install -r requirements.txt for dependencies

Linux Setup

• Creates a Python virtual environment in python-runtime/
• Installs all dependencies from requirements.txt
• Downloads static FFmpeg binary to ffmpeg-bin/

Directory Structure After Setup

After running the setup script, you'll have:

├── python-runtime/          # macOS/Linux Python environment
├── ffmpeg-bin/             # macOS/Linux FFmpeg binary
├── ffmpeg-bin-windows/     # Windows FFmpeg binary
├── audio_extractor.py      # Python script for audio processing
├── requirements.txt        # Python dependencies
└── setup-binaries.sh       # Setup script

Note: Windows builds use system Python, so no python-runtime-windows/ directory is created.

Windows Post-Setup

After running the Windows setup, install Python dependencies using system Python:

pip install -r requirements.txt

No additional setup required - Windows builds use system Python and bundled FFmpeg.

Cross-Platform Builds

• Run ./setup-binaries.sh all to set up binaries for all platforms
• Use the appropriate npm script for your target platform:

npm run build:mac 
npm run build:win 
npm run build:linux

🛠️ Manual Installation Guide

FFmpeg Setup

FFmpeg is required for audio extraction. Choose your platform:

Windows

# Using Chocolatey (recommended)
choco install ffmpeg
# Using Scoop
scoop install ffmpeg
# Manual installation
# Download from [https://ffmpeg.org/download.html](https://ffmpeg.org/download.html)
# Extract to C:\ffmpeg and add C:\ffmpeg\bin to PATH

macOS

# Using Homebrew
brew install ffmpeg
# Manual installation
# Download from [https://ffmpeg.org/download.html](https://ffmpeg.org/download.html)
# Extract to /usr/local/bin/

Linux

# Ubuntu/Debian
sudo apt update && sudo apt install ffmpeg
# CentOS/RHEL
sudo yum install ffmpeg
# Arch Linux
sudo pacman -S ffmpeg

Ollama Setup (Optional)

For AI-powered summaries:

# 1. Download from [https://ollama.ai/](https://ollama.ai/)
# 2. Start Ollama service
ollama serve
# 3. Pull the required model
ollama pull mistral
# 4. Verify installation
ollama list

Verification

Ensure everything is installed correctly:

# Check all dependencies
node --version
python --version
ffmpeg -version
ollama list  # if installed

Setup Troubleshooting

"Python not found" error

• Make sure you've run ./setup-binaries.sh first
• Check that the Python runtime was created successfully
• Verify the paths in transcript.utils.ts match your setup

"FFmpeg not found" error

• Ensure FFmpeg was downloaded by the setup script
• Check that the FFmpeg binary has execute permissions
• Verify the FFmpeg path is correct for your platform

Manual Binary Setup (Alternative)

If the automated setup doesn't work, you can manually set up the binaries:

Python Runtime (macOS/Linux only)

1. Create a virtual environment: python3 -m venv python-runtime
2. Activate it: source python-runtime/bin/activate (macOS/Linux)
3. Install dependencies: pip install -r requirements.txt
4. Deactivate: deactivate

FFmpeg

1. Download from FFmpeg website
2. Extract to ffmpeg-bin/ directory (macOS/Linux) or ffmpeg-bin-windows/ (Windows)
3. Ensure the binary is executable: chmod +x ffmpeg-bin/ffmpeg (macOS/Linux)

Windows Note

Windows builds rely on system Python, so just ensure Python 3.8+ is installed and run:

pip install -r requirements.txt

Python Dependencies

The app requires these Python packages (from requirements.txt):

• openai-whisper>=20231117 - For speech-to-text conversion
• SpeechRecognition>=3.10.0 - Alternative speech recognition
• pydub>=0.25.1 - Audio processing
• torch>=2.0.0 - Machine learning backend
• numpy>=1.24.0 - Numerical computations

Important Notes

Development vs Production

• Development mode: Uses system Python and FFmpeg on all platforms
• Packaged app:
  • macOS/Linux: Uses bundled Python runtime and FFmpeg
  • Windows: Uses system Python + bundled FFmpeg

Gitignore

The binary directories are automatically ignored by git:

• python-runtime/ (macOS/Linux only)
• ffmpeg-bin/ (macOS/Linux)
• ffmpeg-bin-windows/ (Windows) This keeps the repository clean while allowing developers to set up their own binaries.

🎯 Usage

Basic Recording Workflow

1. Launch the application
2. Select recording type (Google Meet, Lesson, or Video)
3. Click "Start Recording"
4. Present your content
5. Click "Stop Recording"
6. Download your video

Advanced Features

Transcript Generation

After recording:

1. Click "Extract Transcript"
2. Wait for processing (audio extraction + speech recognition)
3. Review the transcript and AI-generated summary
4. Files are saved to your desktop automatically

File Organization

Your recordings are automatically organized:

Desktop/
└── captured-videos/
    └── 2024-01-15/
        ├── recording-2024-01-15-14-30-25.webm
        ├── recording-2024-01-15-14-30-25.txt
        └── recording-2024-01-15-14-30-25-summary.txt

⚙️ Configuration

Recording Types

Type	Purpose	Summary Focus
Google Meet	Meeting recordings	Action items, decisions, assignments
Lesson	Educational content	Learning objectives, key concepts, homework
Video	General content	Main points, highlights, insights

AI Model Configuration

Change the AI model in src/main/utils/ollama.utils.ts:

model: 'mistral' // Change to any model you have installed

Speech-to-Text Engines

The app supports multiple STT options:

1. OpenAI Whisper (Local) - Default, works offline
2. Google Speech Recognition (Online) - Fallback option
3. Fallback - Simple text if other options fail

🔧 Development

Development Server

npm run dev

Building for Production

# Windows
npm run build:win
# macOS
npm run build:mac
# Linux
npm run build:linux

🐛 Troubleshooting

Common Issues

FFmpeg Not Found

# Verify installation
ffmpeg -version
# Add to PATH if needed
# Windows: Add C:\ffmpeg\bin to system PATH
# macOS/Linux: Ensure it's in /usr/local/bin/

Ollama Connection Issues

# Start Ollama service
ollama serve
# Check available models
ollama list
# Pull required model
ollama pull mistral
# Test connection
curl http://localhost:11434/api/tags

Python Dependencies

# Install requirements
pip install -r requirements.txt
# Verify audio_extractor.py exists
ls audio_extractor.py

Desktop Sources Not Found

• This is normal on first run
• The app will automatically detect available screens
• Try refreshing or restarting the application

Debug Mode

Run with enhanced logging:

npm run dev -- --debug

Manual Setup

Use the automated setup script:

python setup.py

🏗️ Project Structure

meetingvideo-transrecorder/
├── src/
│   ├── main/           # Electron main process
│   │   ├── handlers/   # IPC handlers
│   │   ├── types/      # TypeScript types
│   │   ├── utils/      # Utility functions
│   │   └── window.ts   # Window management
│   ├── preload/        # Preload scripts
│   │   ├── types/      # Type definitions
│   │   ├── utils/      # API utilities
│   │   └── index.ts    # Main preload
│   └── renderer/       # React frontend
│       ├── components/ # UI components
│       ├── hooks/      # Custom hooks
│       ├── types/      # TypeScript types
│       └── utils/      # Utility functions
├── resources/          # App resources
├── audio_extractor.py  # Python audio processing
├── requirements.txt    # Python dependencies
└── package.json        # Node.js dependencies

🤝 Contributing

We welcome contributions! Here's how to get started:

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Make your changes
4. Test thoroughly
5. Submit a pull request

Development Guidelines

• Follow TypeScript best practices
• Use meaningful commit messages
• Test on multiple platforms
• Update documentation as needed

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆘 Support

Getting Help

1. Check the troubleshooting section above
2. Review console output for error messages
3. Ensure all prerequisites are installed
4. Try running in debug mode
5. Check GitHub Issues

Community

• GitHub Issues: Report bugs and request features
• Discussions: Ask questions and share ideas
• Releases: Download the latest versions

🔄 Continuous Integration

This project uses GitHub Actions for automated builds and releases.

Workflows

Build Artifacts (build-artifacts.yml)

• Triggers: Push to main, pull requests, manual
• Purpose: Build Windows and macOS versions
• Output: Uploads artifacts for manual download

Build and Release (build.yml)

• Triggers: Tags starting with v* (e.g., v1.0.1)
• Purpose: Creates GitHub releases with artifacts
• Output: Automatic release with downloadable files

Using Release Scripts

Linux/macOS:

chmod +x scripts/release.sh
./scripts/release.sh 1.0.1

Windows:

scripts\release.bat 1.0.1

Manual Builds

1. Go to Actions
2. Click "Build Artifacts"
3. Click "Run workflow"
4. Select branch and run

Build Matrix

Platform	Runner	Build Command	Output
Windows	windows-latest	npm run build:win	.exe, win-unpacked/
macOS	macos-latest	npm run build:mac	.dmg, mac/

Artifacts

• Build Artifacts: Available in Actions tab (30-day retention)
• Release Artifacts: Available on Releases page (permanent)

Recent Updates

• ✅ Modular Architecture - Refactored for better maintainability
• ✅ CSS Refactoring - Improved styling with @apply directives
• ✅ GitHub Actions - Automated builds and releases
• ✅ Cross-platform Support - Windows and macOS builds
• ✅ AI Integration - Local AI-powered summarization

---

Made with ❤️ by LinuxDevil

Here's an optimized README structure that maximizes open source appeal, discoverability, and onboarding ease, following best practices and adding some improvements for clarity and search optimization.[^1][^2][^3][^4]

---

Transrecorder – Screen Recorder with Local AI Transcripts & Summaries

Offline, privacy-first screen recorder for Windows, macOS, and Linux. Capture your desktop, save videos locally, and auto-generate transcripts and AI-powered summaries with Ollama—all fully customizable and built with Electron, React, and TypeScript.

[ [ [

---

✨ Features

• 🎥 Screen Recording: Capture your entire screen or specific windows.[^5]
• 📁 Smart File Organization: Automatically saves to organized folders with timestamps.
• 🎵 Audio Extraction: Extract audio from recorded videos for easy processing.
• 📝 AI Transcript Generation: Local, offline speech-to-text via Whisper and more.
• 🤖 AI Summarization: Get context-aware summaries using Ollama with fully offline models.
• 🎯 Recording Profiles: Optimized for meetings (Google Meet), lessons, or general use.
• 👀 Live Preview: View what you're recording in real-time.
• 🔄 Cross-Platform: Seamless support for Windows, macOS, and Linux.
• 🔒 Privacy-First: No data leaves your machine—everything remains local.
• ⚡ Automated Setup: One-command setup for all key dependencies.

---

📸 Screenshots

Main Application Interface

Recording in Progress

---

🚀 Quick Start

Download Pre-built Binaries

1. Go to the Releases page.
2. Download for:
   • Windows: meetingvideo-transrecorder-1.0.0-setup.exe
   • macOS: meetingvideo-transrecorder-1.0.0.dmg
3. Install and launch.

Build from Source

git clone https://github.com/LinuxDevil/MeetingVideo-Transrecorder
cd meetingvideo-transrecorder
npm install
pip install -r requirements.txt
npm run dev

---

📋 Requirements

Component	Version	Purpose
Node.js	16+	Runtime environment
Python	3.8+	Audio processing
FFmpeg	Latest	Video/audio conversion

Optional for AI:

Component	Purpose
Ollama	Local AI summarization engine
Python pkgs	Speech-to-text processing

---

🛠️ Setup Guide

To quickly install everything:

# Platform-specific setup
./setup-binaries.sh
# All platforms
./setup-binaries.sh all
# Or pick one (e.g. macos, windows, linux)

See the Manual Installation Guide below if you prefer to install dependencies yourself.

---

🎯 Usage

Recording Workflow

1. Launch the app.
2. Choose recording mode (Meeting, Lesson, or Video).
3. Click "Start Recording", present content, then "Stop Recording".
4. Download or review your recording and transcript.

Transcript and Summarization

• Click "Extract Transcript" after recording.
• Wait while audio is processed and AI models run.
• Review transcript and the summary (saved automatically).

Organized files will appear as:

Desktop/captured-videos/YYYY-MM-DD/
  ├── recording-2024-09-19-10-10-10.webm
  ├── recording-2024-09-19-10-10-10.txt
  └── recording-2024-09-19-10-10-10-summary.txt

---

🧩 Configuration

• Recording Profiles: Fine-tune for Google Meet, lessons, or generic screen content.
• AI Model: Edit src/main/utils/ollama.utils.ts (model: 'mistral' or any Ollama-supported model).
• Speech-to-Text: Uses Whisper locally (default), Google Speech Recognition (fallback), or plain text if required.

---

🐛 Troubleshooting

• Missing Python/FFmpeg/Ollama: Double-check setup scripts or refer to manual install below.
• Dependencies: Run pip install -r requirements.txt.
• Platform detection issues: App may not detect screens or audio sources first run—restart if needed.
• Run with logs: npm run dev -- --debug

See the FAQ and GitHub Issues for help.

---

📝 Contributing

Contributions welcome—please fork, branch, and PR! See CONTRIBUTING.md for coding style and communication guidelines:

• Use concise, meaningful commit messages.
• Test changes on multiple platforms.
• Update documentation.
• Join GitHub Discussions for questions and ideas.

---

📄 License

Licensed under the MIT License.

---

🙋 Support & Community

• Need help? Open an issue.
• Feature ideas? Join Discussions.
• Stay updated via Releases.

---

Made with love by LinuxDevil

---