修正轉錄內容與指派發言者

You can edit the transcript afterward. If a speaker differs from the automatic detection, reassign it and regenerate the minutes with the confirmed speakers. This page explains editing and speaker assignment.

Last updated: October 1, 2026

Edit the transcript

You can fix the transcript text for each utterance. Correcting misrecognitions and spelling variants also improves the minutes.

1. Open the meeting and select the utterance you want to fix.
2. Use Edit transcript to correct the text and save.
3. Delete an unnecessary utterance with Delete this utterance. Deletion cannot be undone.

Assign speakers

Automatically detected speakers get labels such as You, Other, or Speaker 1. You can assign names from the meeting participants or contacts to label the correct speaker.

Use Play voice to check a speaker’s utterances before assigning, which reduces mix-ups.

1. Open Speaker assignment.
2. Select a speaker and specify a person from “Participants in this meeting” or “Select from contacts”.
3. If needed, use Play voice to check the utterances.
4. If unassigned speakers remain, use Estimate remaining speakers to match voices and estimate them together.

Rename or merge speakers

When renaming a speaker, choose whether to apply it to only that block or to all.

If one person’s utterances are split across speakers, you can merge them with Set as you. To return to the automatic detection, use Restore detected speaker.

Confirm and regenerate the minutes

Once you confirm the speaker assignment, you can regenerate the minutes with that speaker information.

If unassigned speakers remain, the count is shown before confirming. If you continue with them unassigned, they appear in the minutes as labels such as “Speaker 1”.

1. Review the assignments.
2. Press Confirm and regenerate minutes.
3. The minutes are updated to reflect the speaker information.