Rename speakers in a transcript

You can rename any speaker directly from the transcript. The name applies to every line from that speaker in the recording — not only the line you started from — and everyone with access to the recording sees it.

On iOS

1. Open the recording and switch to the transcript.
2. Tap the speaker's name above any of their lines (it has a small pencil next to it).
3. In the Speaker name dialog, type the name and tap Save.

The dialog notes exactly what happens: "Name this speaker. Applies to every line they speak."

On Android

1. Open the recording and switch to the transcript.
2. Tap the speaker's name above any of their lines.
3. In the Speaker name dialog, type the name and save.

On the web

1. Open the recording and click a speaker's name in the transcript.
2. Choose Rename speaker.
3. Type the name and confirm. The dialog states: "The name applies to every line from this speaker, for everyone with access."

Who sees the new name?

Everyone with access to the recording — you on all your devices, members of any Space the recording is shared in, and viewers of a public share link. The name is also used everywhere the transcript is rendered: exports, copied text, and AI answers that quote the transcript. There is no per-viewer name.

Renaming requires edit access: the owner and Space members can rename speakers; someone viewing through a public link cannot. See Who can see and do what?

How do I undo a rename?

Reset the speaker to its automatic name:

• iOS / Android — tap the speaker's name again and choose Reset to auto in the Speaker name dialog.
• Web — open the speaker's menu and choose Reset to auto name.

Resetting restores whatever Canto had before your manual name: the AI-inferred name if there is one, otherwise the numbered fallback like Speaker 1. See How does Canto tell speakers apart? for where those automatic names come from.

Troubleshooting

The rename changed lines that belong to a different person

Names attach to a detected speaker, not to lines. If Canto merged two people into one speaker during transcription, renaming that speaker labels both people's lines — and there's no way to reassign individual lines. See Speakers are mixed up or mislabeled.

The same person appears twice under two names

The opposite case: one voice was split into two speakers. Rename both labels to the same name; the lines stay separate turns but read correctly.