Transcribe

Turn recordings and your own voice into text, privately.

Before you start

Transcription uses Whisper. The first time you open Transcribe, pick a model to download:

• Whisper Base (recommended to start): quick, and great for dictation.
• Whisper Small: more accurate for files.
• Whisper Large v3 Turbo: the most accurate, for important recordings.

You can install more than one. When transcribing a file, choose which to use from the Model menu.

Transcribe a file

1. Choose Transcribe a file.
2. Optionally pick the language. Detect automatically works well in most cases.
3. Drop an audio or video file onto the box, or click to choose one: MP3, M4A, WAV, MP4, MOV and more, up to 2 GB.
4. Watch the progress. You can leave the page, and it continues in Activity.

Live transcription

1. Choose Live and click Start listening. Allow microphone access the first time.
2. Speak. Text appears as you talk: grey words are still settling, black words are final.
3. Click Stop. The transcript is saved to the list.

Read, fix and export

Pick a transcript from the list on the left:

• Click a timestamp to hear that moment. The sentence being played is highlighted.
• Double-click a sentence to correct it. Press Enter to save or Esc to cancel.
• Click the title to rename it.
• TXT · SRT · VTT · JSON export plain text, subtitles or data.
• 🗑 then Confirm deletes it.

Search the list to find any transcript by title or words.

Dictate into any app

See Dictation.

Transcribe from a link

Paste a direct link to an audio or video file into …or paste a direct link under the drop area and click Transcribe. VoxStudio downloads it and transcribes it as if you'd dropped the file. Web pages and links to your local network aren't accepted.