generated: '2026-09-16'
method: derived
source: Derived from components.schemas $ref links and id-reference fields in openapi/pyannoteai-api-api-openapi.yml,
  openapi/pyannoteai-media-api-openapi.yml and openapi/pyannoteai-streaming-api-openapi.yml.
docs: https://docs.pyannote.ai/api-reference/schemas/diarizationschema
notation: relationships use has_one / has_many / belongs_to with the field that carries the link; direction
  is from the owning entity.
entities:
- name: Team
  id_prefix: null
  domain: account
  description: The account scope an API key belongs to; GET /v2/jobs lists a team's jobs. Not a schema
    in the spec.
- name: Job
  id_prefix: null
  domain: jobs
  schemas:
  - JobCreated
  - JobListItem
  description: An asynchronous unit of work identified by jobId, with JobStatus and createdAt/updatedAt.
- name: DiarizationJob
  domain: jobs
  schemas:
  - DiarizationJob
  - DiarizationJobOutput
  - DiarizationSegment
  description: A job created by POST /v1/diarize; output holds diarization segments and optional transcription.
- name: VoiceprintJob
  domain: voiceprints
  schemas:
  - VoiceprintJob
  - VoiceprintJobResults
  description: A job created by POST /v1/voiceprint; output.voiceprint is the extracted voiceprint string.
- name: IdentifyJob
  domain: identification
  schemas:
  - IdentifyJob
  - IdentificationJobOutput
  - IdentificationSegment
  - IdentificationVoiceprint
  description: A job created by POST /v1/identify; matches diarized speakers to supplied voiceprints.
- name: Voiceprint
  domain: voiceprints
  schemas:
  - Voiceprint
  description: A {label, voiceprint} pair supplied in IdentifyRequest.voiceprints. Voiceprints are not
    stored server-side as addressable resources; the caller keeps them.
- name: Media
  domain: media
  schemas:
  - GetMediaUploadUrl
  - MediaResponse
  description: A media:// temporary storage object referenced by url in job requests.
- name: Stream
  domain: streaming
  schemas:
  - Stream
  - StreamCreated
  - CreateStream
  - StreamStatus
  description: A live diarization stream (id, status, startedAt, completedAt); StreamCreated returns the
    WebSocket url.
relationships:
- from: Team
  to: Job
  kind: has_many
  via: GET /v2/jobs items[].id
- from: DiarizationJob
  to: Job
  kind: belongs_to
  via: jobId
  note: GET /v1/jobs/{jobId} returns oneOf DiarizationJob | VoiceprintJob | IdentifyJob
- from: VoiceprintJob
  to: Job
  kind: belongs_to
  via: jobId
- from: IdentifyJob
  to: Job
  kind: belongs_to
  via: jobId
- from: VoiceprintJob
  to: Voiceprint
  kind: has_one
  via: output.voiceprint
- from: IdentifyJob
  to: Voiceprint
  kind: has_many
  via: input voiceprints[].voiceprint (IdentifyRequest)
- from: IdentifyJob
  to: Voiceprint
  kind: has_many
  via: output IdentificationVoiceprint.match (label)
- from: DiarizationJob
  to: Media
  kind: belongs_to
  via: url (DiarizeRequest; may be a media:// url)
- from: IdentifyJob
  to: Media
  kind: belongs_to
  via: url (IdentifyRequest)
- from: VoiceprintJob
  to: Media
  kind: belongs_to
  via: url (VoiceprintRequest)