FAQ
Common questions about how Verso works.
Are my videos private?
Yes. Verso processes all video files locally on your device. Your videos are never uploaded to any server. Only vocabulary metadata — words, translations, timestamps, and review scores — is synced to the cloud.
What video formats does Verso support?
Verso supports any format that your operating system can play, including MP4, MKV, AVI, MOV, and WebM. If you can open it in a media player, Verso can process it.
How accurate is the transcription?
Transcription quality depends on the Whisper model you use and the audio quality of the video. The base model works well for clear speech in most major languages. For better accuracy, use the medium or large model — though these are slower to run.
Can I import subtitles instead of transcribing?
Not in the current version. Subtitle import is planned for a future release. For now, Verso transcribes all audio using WhisperX.
Does Verso work offline?
Yes, for all core features. Transcription uses WhisperX (local). The default translation provider is Argos Translate (local). Flashcard review works offline. Sync requires an internet connection, but all local data is preserved if you are offline.
Which languages are supported?
WhisperX supports over 100 languages, including Spanish, French, German, Japanese, Mandarin, Korean, Portuguese, Arabic, Hindi, Russian, and many more. Translation support depends on the provider you choose.
Can I use the mobile app without a desktop?
You can review existing flashcards on mobile at any time. However, processing new videos requires the desktop app — mobile devices do not have enough resources to run WhisperX locally.
How is Verso different from Language Reactor?
Language Reactor works with streaming video in the browser and requires an internet connection. Verso works with video files you own locally, runs entirely offline, and focuses on spaced repetition review rather than in-player lookup.
Does Verso store my videos?
No. Verso never stores or uploads video files. The server only ever receives vocabulary metadata. Video files stay on the device where you processed them.
Have a question not answered here? Open an issue on GitHub →