Diarization

Speaker diarization partitions transcribed speech by speaker - answering “who spoke when”. When enabled, each piece of transcribed text is labelled with a speaker_id so you can tell participants apart in a conversation.

Diarization is available on the Prerecorded and Realtime APIs. It is not available on the Turns API.

Enabling Diarization

Set the diarize query parameter to true. Optionally cap the number of distinct speakers with max_speakers (1-4); leave it unset to let the server determine the count automatically.

?diarize=true&max_speakers=2

Speaker IDs

Each speaker is identified by an integer speaker_id, starting at 0. IDs are assigned in the order speakers are first heard and remain stable for the duration of a session.

Anonymous labels

Speaker IDs are anonymous labels, not identities. speaker_id: 0 is simply “the first speaker heard” - diarization does not recognise who a person is across sessions.

In the Response

On prerecorded, the response is split into per-speaker segments, each with its own speaker_id and text. On realtime, each transcript message carries the speaker_id of its dominant speaker. See the API reference for Prerecorded and Realtime for the full response fields.