Vidge Desktop
AI-Powered Video Management That Runs Entirely On Your Computer
Your videos. Your hardware. Your privacy.
Vidge Desktop processes everything locally - no cloud subscriptions, no upload queues, no privacy concerns.
Why Vidge Desktop?
Process Videos Instantly
Import videos and watch them get processed in real-time. Transcription, face detection, and scene analysis happen on your machine - no upload queues.
Privacy First
Your videos never leave your computer. Perfect for legal recordings, medical videos, HR materials, and confidential business footage.
No Subscriptions Required
Unlike cloud services that charge monthly fees + per-minute processing costs, process unlimited videos with no ongoing fees.
Find Anything Fast
Search across transcripts, people, tags, and metadata. Find that exact moment someone said something in seconds.
Core Features
VidgeScribe - AI Transcription
Local speech-to-text powered by OpenAI's Whisper model:
- Multiple Models - Choose from Tiny (fast) to Large (accurate)
- Word-Level Timestamps - Click any word to jump to that moment
- 99+ Languages - Automatic language detection
- Unlimited Processing - No per-minute fees
- Export Options - SRT subtitles or plain text
Face Recognition & People Tracking
- Automatic Detection - Finds faces in every video
- Cross-Video Matching - Same person recognized across your library
- People Database - Grows smarter over time
- Drag-to-Merge - Combine duplicate people easily
- Appearance Timeline - See when each person appears
- Filter by Person - Find all videos featuring someone
Smart Tagging System
- Manual Tags - Right-click any video to add tags
- Timecoded Tags - Bookmark specific moments with timestamps
- Auto-Tags - Generated from folder names and filenames
- Tag Management - Dedicated page to view, rename, merge, and delete tags
- Click to Jump - Timecoded tags navigate directly to that moment
Scene Detection
- Automatic Detection - Finds scene changes in your videos
- Visual Timeline - Thumbnail strip of all scenes
- Click to Jump - Navigate to any scene instantly
- Perfect for Long Recordings - Quickly scrub through hours of footage
Full-Text Search
- Search Everything - Transcripts, tags, people, metadata
- Real-Time Results - Results appear as you type
- Smart Ranking - Most relevant results first
Perfect For
Content Creators
Organize footage from multiple shoots, find specific moments instantly, build a searchable content library.
Legal & Compliance
Process depositions locally, keep sensitive footage off the cloud, searchable transcripts for case prep.
Education & Training
Manage lecture recordings, create searchable video libraries, tag by course or instructor.
Corporate Communications
Process meeting recordings privately, find who said what when, build an internal video knowledge base.
Getting Started
- 1
Download & Install
Download Vidge Desktop from your account portal and install on Windows 10/11.
- 2
Import Videos
Drag-and-drop videos onto the window or click "Add Videos" in the header.
- 3
AI Processing
Watch as Vidge automatically transcribes, detects faces, and analyzes scenes.
- 4
Search & Organize
Use the search bar to find anything, add tags, and browse by people.
Comparison vs Cloud Services
| Feature | Vidge Desktop | Cloud Services |
|---|---|---|
| Transcription | Unlimited, local | Pay per minute |
| Face Recognition | Unlimited, local | Pay per analysis |
| Privacy | 100% local | Uploaded to cloud |
| Processing Speed | Instant start | Upload + queue |
| Monthly Cost | None | $20-100+/month |
| Offline Use | Full functionality | Requires internet |
System Requirements
Minimum
- Windows 10/11 (64-bit)
- 8GB RAM
- 500MB disk space
- Any modern CPU
Recommended
- Windows 11 (64-bit)
- 16GB RAM
- SSD storage
- NVIDIA GPU (faster AI)
AI Models
Models are automatically downloaded on first use:
| Model | Size | Purpose |
|---|---|---|
| Whisper Base | ~142 MB | Speech-to-text (default) |
| Whisper Large | ~3 GB | High-accuracy transcription |
| FaceONNX | ~100 MB | Face detection & recognition |
Storage Structure
Your Library Location/ ├── videos/ │ └── My_Video_Title/ │ ├── video-record.json # Metadata & transcripts │ ├── thumbnails/ # Preview images │ ├── scenes/ # Scene detection frames │ └── faces/ # Detected face crops ├── persons/ # Face recognition database ├── models/ # AI models (auto-downloaded) ├── tags.json # Tag definitions └── logs/ # Debug logs
Troubleshooting
Transcription is Slow
- • Try a smaller model (Tiny or Base) in Settings
- • NVIDIA GPU significantly speeds up processing
- • Close other resource-intensive applications
Face Recognition Not Finding People
- • Faces must be at least 60px in the video
- • Check video quality (low resolution = poor detection)
- • Ensure good lighting in the footage
Import Fails
- • Supported formats: MP4, MOV, AVI, MKV, WebM
- • Check disk space for thumbnails and models
- • Try re-encoding with HandBrake if file is corrupted
Need Help?
Check the FAQ or contact support@ltclabs.dev