Transcribe audio in every language

MacOS: install .dmg file from releases

Windows: install .exe file from releases

Linux: install either .deb (recommend) or use appImage from releases and follow

MacOS Windows Linux

• Transcribe almost every language
• Works offline, no data sent out
• Easy to use
• Model parametrs for advanced users
• Optimized for CPU (Windows / Linux)
• Optmized for GPU (MacOS)
• Support SRT, VTT formats
• Transcribe audio / video

Windows: Version 8 greater.

Hardware: No special requirement. resource usage can be customized through advanced settings in main window.

Currenly, MacOS computers get the best performance since there's GPU optimizations.

Please install the following packages

sudo apt-get install ffmpeg libgcc-s1 libc6 zlib1g libselinux1 libpcre3 libbz2-1.0 liblzma5 libcap2 libexpat1 libgpg-error0 libdbus-1-3 libcom-err2 libc6 libncursesw6 libtinfo6 libpulse0 libc6 libkeyutils1 libc6

And for install the deb file -> download it and execute

dpkg -i ./vibe*.deb

Currenly, listening for the audio file isn't supported on Linux And using the appImage isn't recommended.

1. Stop the program

2. Download it from ggml-medium.bin

3. Place it in the folder C:\Users\User\AppData\Local\github.com.thewh1teagle.vibe With the exact name ggml-medium.bin

4. Start the program

And if the antivirus still block the model file, you can always exclude it from the antivirus.

PRs are welcomed! In addition, you're welcome to add translations.

You can see the roadmap in Vibe-Roadmap

1. Copy en.json from locales folder
2. Create new file eg ru.json
3. Change every value to the new language and keep the keys as is
4. create PR / issue in github

see BUILDING.md

Medium post