Speaker labels on device

Offline Speaker Diarization: See Who Said What

Speaker diarization answers a different question than transcription. Transcription tells you what was said; diarization separates the conversation into who spoke when.

Readable conversations instead of a wall of text

SonWeave groups consecutive turns, displays distinct speaker colors and lets you replace generic labels with real names. That makes interviews and meetings easier to scan and quote.

  • Automatic speaker turns
  • Color-coded labels
  • Rename speakers
  • Tap text to revisit audio

Private diarization

Speaker analysis runs locally. SonWeave does not need to upload the meeting to its own server to separate voices, which matters when the identity of participants is itself sensitive.

Exports that preserve context

Supported text and subtitle exports can include speaker labels and timestamps, so the useful structure is not trapped inside the app.

FAQ

Frequently asked questions

Is speaker diarization the same as voice recognition?

No. Diarization separates distinct speakers; it does not automatically identify a person's legal identity.

Can I rename speakers?

Yes. Replace generic speaker labels with names that make sense for the conversation.

Does diarization work offline?

Yes. SonWeave's speaker-label workflow is designed to run on the device.