Retranscription¶
A Second Reading¶
Hidden Feature
Not satisfied with the transcript? You can re-run transcription with different settings — no need to re-record.
When to Retranscribe¶
- The transcript has errors and you want to try a larger model
- You initially used the wrong language setting
- The speaker count was off and diarization suffered
- You want to try a different combination of settings
How to Retranscribe¶

- Select a different Whisper model
- Optionally change the language
- Optionally adjust the speaker count
- Click Retranscribe
This Replaces the Existing Transcript
Retranscription clears the current transcript and speaker assignments, then generates new ones. Summaries and images are not affected.
Progress Tracking¶
You'll see live progress updates through the same phases as the original:
- Transcription progress with estimated time remaining
- Speaker identification
- Session naming (if the name was still generic)
Next: Assign Speaker Names →