Release Notes
Support for Audio File Transcription
Upload audio files directly in Assistant or Vault and Harvey will automatically generate a transcript you can read, edit, and query.
What’s New
Upload an audio file directly in an Assistant conversation or to Vault, and Harvey transcribes it automatically. The transcript is returned as a Word (.docx) file you can read, query over, and edit. Key features include:
- Automatic speaker labels (”Speaker 1”, “Speaker 2”) and a timestamp at the start of each speaker's turn
- Automatic language detection, so you don't need to specify the language
- Support for M4A, MP3, WAV, WebM, FLAC, and OGG files, up to 2 hours per file
- Two ways to correct a transcript: the Word editor side panel or a request to Assistant (for example, renaming a speaker label)
Why It Matters
You no longer need a separate tool or workflow to turn your audio files from depositions, client interviews, hearings, and calls into usable text. This feature brings transcription directly into Harvey, allowing you to transform a raw recording into a searchable, editable transcript without leaving the web app.
How To Use
For instructions and tips on using audio transcription, refer to Transcribing Audio Files in Harvey.