Tips & Best Practices

Improve Transcription Accuracy

1. Use better models: medium or large models are significantly more accurate than base
2. Specify language: Auto-detect is less accurate than specifying
3. Enable stable timestamps: VAD mode gives more precise timing
4. Audio preprocessing: Pro users can use audio enhancement to reduce noise
5. Avoid overlapping speech: Multiple speakers talking simultaneously reduces accuracy

Efficient Workflows

Meeting Notes

1. Enable “Live recording transcription” (Pro)
2. Capture both microphone and system audio
3. Enable “Speaker diarization” (Pro)
4. Generate AI summary after transcription
5. Export as DOCX to share with team

Podcast Subtitles

1. Import audio file
2. Transcribe with medium model
3. Enable stable timestamps
4. Fine-tune timestamps if needed
5. Export SRT or burn subtitles directly (Pro)

Video Batch Processing

1. Set up “Watch folder” (Pro)
2. Drop downloaded videos into the folder
3. Auto-transcribe and export SRT
4. Use subtitle burn-in for finished product (Pro)

Shortcut Cheatsheet

Custom Vocabulary

In Settings → Vocabulary, add professional terms, names, brands, etc. The transcriber will prioritize matching these terms, improving accuracy in professional scenarios.