Guide

Identify speakers in a transcript

Request Pro speaker diarization and consume optional speaker values in completed segments.

Built around the real API

Add speaker structure without changing the job lifecycle

Diarization is an optional job flag. Creation and polling stay the same; completed segments can include numeric or string speaker identifiers when the processing result provides them.

One boolean

Set speaker_diarization to true on the transcription request.

Plan checked

The API returns pro_required when the account cannot use the feature.

Segment based

Speaker values belong to timestamped segments rather than the full text string.

Optional output

Clients should tolerate segments without a speaker value.

How it works

From request to transcript

A predictable sequence you can reuse in server applications, automations, and internal tools.

  1. 01

    Enable the flag

    Set speaker_diarization: true for a URL or uploaded-file job.

  2. 02

    Poll normally

    Use the same Location header and job endpoint.

  3. 03

    Render segment speakers

    Group or label segments using speaker when the field is present.

Copy and adapt

Make the request

Examples use the production API origin and an environment variable for the secret key.

Request
curl --include --request POST https://fast-transcriber.com/api/v1/transcriptions \
  --header "Authorization: Bearer $FAST_TRANSCRIBER_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
    "source_url": "https://cdn.example.com/media/panel-discussion.mp3",
    "speaker_diarization": true
  }'
Keep credentials server-side
Never expose a Fast Transcriber API key in public browser JavaScript, a mobile bundle, source control, or a prompt.
Before you ship

Requirements and boundaries

  • Speaker diarization requires a Pro account.
  • The flag is false by default.
  • Read speaker identifiers from individual segments and format display labels in your application.
  • Keep a fallback for results where a segment has no speaker value.
Common questions

Frequently asked questions

What happens on a non-Pro account?+

The request returns 403 with the pro_required error code.

Is speaker a top-level response field?+

No. It can appear on individual timestamped segments.

Does enabling diarization change polling?+

No. The job remains asynchronous and uses the same status endpoint.

© 2026 FastTranscriberBuilt for developers who work with speech.