Skip to main contentTranscribing Interviews
Transcribe audio and video files into text for analysis.
Settings
Configure default transcription settings in Workspace Settings > Transcription Settings:
- Auto-transcribe on upload: Start transcription automatically
- Model selection: Choose the transcription model
- Language selection: Set default language or auto-detect
- Redaction: Enable automatic redaction of sensitive information
- Filler words: Include or exclude filler words
- Custom vocabulary: Add unique words, names, or jargon (one word per line)
Settings apply to new uploads by default but can be overridden per transcription.
Process
- Create a new document
- Click Upload audio or video in the document editor
- Select your file
- Click Transcribe and configure settings:
- Redaction: Redact sensitive information (names, addresses)
- Exclude filler words: Remove “uh,” “um,” etc.
- Language: Select language or use Autoselect
- Video: .mp4, .webm, .ogv, .avi, .mov, .mkv
- Audio: .mp3, .wav, .ogg, .aac, .webm, .flac
Models
General Purpose
- Nova-2: Latest model, supports all available languages
- Enhanced General Purpose: Improved accuracy, subset of languages
- Base General Purpose: Moderate language support
Specialized (English only)
- Meeting: Multi-speaker meetings
- Phone Call: Two-speaker conversations
- Financial: Financial terminology
- Medical: Healthcare terminology
- Automotive: In-car audio
Specialized models currently support English only.
Languages
Nova-2 supports: Bulgarian, Catalan, Chinese (Simplified and Traditional), Chinese (Cantonese), Czech, Danish, Dutch, English (US, Australia, UK, New Zealand, India), Estonian, Finnish, Flemish, French, French (Canada), German, German (Switzerland), Greek, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Latvian, Lithuanian, Malay, Norwegian, Polish, Portuguese (Brazil, Portugal), Romanian, Russian, Slovak, Spanish, Spanish (Latin America), Swedish, Thai, Turkish, Ukrainian, Vietnamese. 