Your privacy matters
We built Transcribe with privacy as a core principle. Here's exactly what we do and don't do with your data.
The short version
What we do
- ✓ Process everything in your browser
- ✓ Keep local files on your device
- ✓ Cache public video transcripts
- ✓ Use privacy-friendly analytics
What we don't do
- ✗ Upload your local files
- ✗ Track you across websites
- ✗ Store personal information
- ✗ Use invasive cookies
How Transcribe works
Local files
Files you select from your device are processed entirely in your browser. They are never uploaded to our servers. The file data, transcription, and all processing stay on your device. We don't even know what files you're transcribing.
Online videos
Videos from Twitter/X and YouTube are streamed through our server due to browser security restrictions. We don't store the videos - they're streamed directly to your browser for processing.
AI transcription
All transcription happens locally in your browser using WebAssembly. The audio never leaves your device during processing. We use OpenAI's Whisper model running entirely client-side.
Community caching
Transcriptions from public URLs (not local files) are saved to a community cache. This helps other users get instant results for the same videos. Cached data includes: video URL, transcription text, and model used. Local file transcriptions are never cached.
Analytics and tracking
We use Umami Analytics, a privacy-focused, cookie-free analytics service. It helps us understand how the service is used without tracking you personally. We respect your browser's Do Not Track setting.
What we collect
- • Page views and basic usage patterns
- • Transcription events (success/failure rates)
- • Source type (local/YouTube/Twitter/URL)
- • Whether the Whisper Large v3 Turbo model was reused from local browser cache
- • Error categories (no error details)
- • Processing duration (aggregate only)
- • Browser type and device information
- • General location (country level only)
What we don't collect
- ✗ IP addresses or personal identifiers
- ✗ Cookies or tracking pixels
- ✗ Specific video URLs or titles
- ✗ Transcription content or text
- ✗ File names or paths
- ✗ Individual user tracking
What we store
On your device
- • Whisper Large v3 Turbo model files (browser cache)
- • Whether the model is cached
This data stays on your device and can be cleared anytime.
In our database
- • Video URLs
- • Transcription text and segments
- • Model identifier used
- • Detected language
- • Processing timestamp
Only for public videos, never local files.
Questions about privacy?
We're committed to transparency. If you have any questions about how we handle your data, we're here to help.
Get in touchThis privacy policy is written in plain language because privacy policies should be understandable by everyone.