Questions
Which languages does it support?
English, Hindi, Urdu, Arabic, Bengali, Marathi, Tamil, Telugu and around 90 others, using OpenAI’s open-source Whisper model. Choose “Accurate” for the best results in Hindi and Urdu.
Is my audio uploaded?
No. The speech recognition runs inside your browser. Your recording never leaves your device.
Can it make subtitles for my video?
Yes. Drop in an MP4 or WebM video and download an SRT file you can upload to YouTube or open in any video editor.
How long can the file be?
Up to 60 minutes. A typical computer handles a few minutes of audio per minute; phones are slower. The first run also downloads the model once.
Can it translate?
Yes. Set Output to “Translate to English” and speech in any supported language comes out as English text.