AitJaz

Questions from the bench

Which languages are supported?

AitJaz supports 18 languages for speech recognition. English and 13 more languages (Spanish, German, French, Portuguese, Turkish, Chinese, Japanese, Arabic, Hindi and others) run on a universal multilingual model that works best in English. Russian uses a dedicated model, while Kazakh, Uzbek and Kyrgyz use a fast model made specifically for these languages. The app interface is available in 15 languages.

Where does my voice go?

By default, your voice data stays entirely on your computer and is never sent anywhere. AitJaz processes dictation audio on your PC and discards it. If you choose to enable cloud recognition by providing your own API key for services like OpenAI or Google Gemini, audio will go to that service. The company only receives your data if you opt-in to send a problem report or take a recognition test.

Do I need internet?

You only need internet to download the app and speech model initially. After that, everything works offline unless you choose to enable cloud recognition yourself. For speaker separation in meetings, you need to download a 27 MB voice model once.

Will it run on my computer?

Most likely yes. AitJaz does not require a graphics card. On first launch, the app checks your PC and shows how fast it will run, allowing you to test this before purchasing.

Can it transcribe audio and video files?

Yes. You can open Audio & video to text in settings, select a recording, and save the text as Word, .txt or .srt subtitle files. The app can label speakers if you enable speaker separation. It supports the same 18 languages as dictation and can also record Zoom, Teams or Google Meet calls directly.