Made for creators
Live captions and an OBS overlay, built around your existing stream tools.
Free and open source
No fees, no subscription, no API key required.
Local processing
Data never leaves your computer. All audio and text processing happens where you have full control over it.
Scrybe runs speech recognition on your computer. Choose from one of the built-in presets, or use your own model file.
Speaker-aware transcripts
Further improvements to caption segmentation and speaker identification are planned.
From installation to captions on stream.
Install the desktop app, prepare your captions, and add the overlay to an OBS scene.
-
Install and launch Scrybe
Download Scrybe for your operating system. The app currently downloads a model preset for local transcription.
-
Style your captions
Open the Overlay page to adjust the caption box’s position, size, colors, and text.
-
Add the overlay to OBS
Create a browser source in OBS and paste this local overlay URL.
http://localhost:3030/app/v1/overlay -
Start captions
Choose your audio source in Settings, then select Start captions. Your captions appear in the overlay.
Powered by whisper.cpp.
Scrybe embeds whisper.cpp for local speech recognition. Model files are downloaded for use on your computer.
whisper.cpp is Georgi Gerganov's port of OpenAI's Whisper speech-recognition model to plain C/C++. More at whisper.cpp.
- Native application
- Rust/C++
- Hardware acceleration depends on the platform and build
- CPU + GPU
- No transcription API subscription
- $0