asyncapi: 3.0.0
info:
  title: Gladia real-time WebSocket
  version: 1.0.0
  description: >-
    Use the Gladia WebSocket to stream audio and receive real-time
    transcription,

    events, add-on results, and post-processing outputs.
servers:
  real-time:
    host: api.gladia.io/v2/live?token={SESSION}
    protocol: wss
    description: |-
      Connect using the session token returned by POST /v2/live.
      The same URL can be reused after recoverable disconnects.
channels:
  liveTranscription:
    address: liveTranscription
    description: >-
      Single WebSocket channel for all client->server and server->client
      messages.
    messages:
      audioChunkJson:
        $ref: '#/components/messages/audioChunkJson'
      audioChunkBinary:
        $ref: '#/components/messages/audioChunkBinary'
      stopRecording:
        $ref: '#/components/messages/stopRecording'
      transcript:
        $ref: '#/components/messages/transcript'
      speechStart:
        $ref: '#/components/messages/speechStart'
      speechEnd:
        $ref: '#/components/messages/speechEnd'
      translation:
        $ref: '#/components/messages/translation'
      namedEntityRecognition:
        $ref: '#/components/messages/namedEntityRecognition'
      sentimentAnalysis:
        $ref: '#/components/messages/sentimentAnalysis'
      postTranscript:
        $ref: '#/components/messages/postTranscript'
      postFinalTranscript:
        $ref: '#/components/messages/postFinalTranscript'
      postChapterization:
        $ref: '#/components/messages/postChapterization'
      postSummarization:
        $ref: '#/components/messages/postSummarization'
      audioChunkAck:
        $ref: '#/components/messages/audioChunkAck'
      stopRecordingAck:
        $ref: '#/components/messages/stopRecordingAck'
      startSession:
        $ref: '#/components/messages/startSession'
      startRecording:
        $ref: '#/components/messages/startRecording'
      endRecording:
        $ref: '#/components/messages/endRecording'
      endSession:
        $ref: '#/components/messages/endSession'
operations:
  sendAudioChunkJson:
    action: receive
    channel:
      $ref: '#/channels/liveTranscription'
    summary: Send an audio chunk as JSON (base64).
    messages:
      - $ref: '#/channels/liveTranscription/messages/audioChunkJson'
  sendAudioChunkBinary:
    action: receive
    channel:
      $ref: '#/channels/liveTranscription'
    summary: Send an audio chunk as a binary frame.
    messages:
      - $ref: '#/channels/liveTranscription/messages/audioChunkBinary'
  sendStopRecording:
    action: receive
    channel:
      $ref: '#/channels/liveTranscription'
    summary: Stop recording.
    messages:
      - $ref: '#/channels/liveTranscription/messages/stopRecording'
  onTranscript:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/transcript'
  onSpeechStart:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/speechStart'
  onSpeechEnd:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/speechEnd'
  onTranslation:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/translation'
  onNamedEntityRecognition:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/namedEntityRecognition'
  onSentimentAnalysis:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/sentimentAnalysis'
  onPostTranscript:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/postTranscript'
  onPostFinalTranscript:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/postFinalTranscript'
  onPostChapterization:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/postChapterization'
  onPostSummarization:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/postSummarization'
  onAudioChunkAck:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/audioChunkAck'
  onStopRecordingAck:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/stopRecordingAck'
  onStartSession:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/startSession'
  onStartRecording:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/startRecording'
  onEndRecording:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/endRecording'
  onEndSession:
    action: send
    channel:
      $ref: '#/channels/liveTranscription'
    messages:
      - $ref: '#/channels/liveTranscription/messages/endSession'
components:
  messages:
    audioChunkJson:
      name: audio_chunk
      title: Audio Chunk (JSON)
      contentType: application/json
      summary: Send audio bytes as base64-encoded chunk.
      payload:
        type: object
        required:
          - type
          - data
        properties:
          type:
            type: string
            const: audio_chunk
          data:
            type: object
            required:
              - chunk
            properties:
              chunk:
                type: string
                description: Base64-encoded raw audio bytes.
      description: >-
        Client payload to stream audio in JSON when binary frames are not
        convenient.
      examples:
        - payload:
            type: audio_chunk
            data:
              chunk: UklGRiQAAABXQVZFZm10IBIAAAABAAEAESsAACJWAAACABAAZGF0YQAAAA==
    audioChunkBinary:
      name: audio_chunk_binary
      title: Audio Chunk (Binary frame)
      contentType: application/octet-stream
      summary: Send audio bytes as a binary WebSocket frame.
      payload:
        type: string
        format: binary
        description: >-
          Raw audio bytes matching the session
          encoding/bit_depth/sample_rate/channels.
      description: Client streams raw audio as binary frames for lowest overhead transport.
    stopRecording:
      name: stop_recording
      title: Stop Recording
      contentType: application/json
      summary: Inform Gladia that recording is over.
      payload:
        type: object
        required:
          - type
        properties:
          type:
            type: string
            const: stop_recording
      description: >-
        Client command indicating no more audio frames will be sent for this
        session.
      examples:
        - payload:
            type: stop_recording
    transcript:
      name: transcript
      title: Transcript
      contentType: application/json
      summary: >-
        This message contains the informations about the transcript of the audio
        chunk.
      payload:
        $ref: '#/components/schemas/TranscriptMessage'
    speechStart:
      name: speech_start
      title: Speech Start
      contentType: application/json
      summary: Indicates the start of detected speech activity on a given channel.
      payload:
        $ref: '#/components/schemas/SpeechEventMessage'
    speechEnd:
      name: speech_end
      title: Speech End
      contentType: application/json
      summary: Indicates the end of detected speech activity.
      payload:
        $ref: '#/components/schemas/SpeechEventMessageEnd'
    translation:
      name: translation
      title: Translation
      contentType: application/json
      summary: >-
        This message contains the informations about the translation of the
        audio chunk.
      payload:
        $ref: '#/components/schemas/TranslationMessage'
    namedEntityRecognition:
      name: named_entity_recognition
      title: Named Entity Recognition
      contentType: application/json
      summary: >-
        This message contains the informations about the named entity
        recognition of the audio chunk.
      payload:
        $ref: '#/components/schemas/NamedEntityRecognitionMessage'
    sentimentAnalysis:
      name: sentiment_analysis
      title: Sentiment Analysis
      contentType: application/json
      summary: >-
        This message contains the informations about the sentiment analysis of
        the audio chunk.
      payload:
        $ref: '#/components/schemas/SentimentAnalysisMessage'
    postTranscript:
      name: post_transcript
      title: Post Transcript
      contentType: application/json
      summary: >-
        This message contains the informations about the post transcript of the
        audio chunk.
      payload:
        $ref: '#/components/schemas/PostTranscriptMessage'
    postFinalTranscript:
      name: post_final_transcript
      title: Final transcript (aggregated)
      contentType: application/json
      summary: >-
        This message contains the informations about the post final transcript
        of the audio chunk.
      payload:
        $ref: '#/components/schemas/PostFinalTranscriptMessage'
    postChapterization:
      name: post_chapterization
      title: Chapterization
      contentType: application/json
      summary: >-
        This message contains the informations about the post chapterization of
        the audio chunk.
      payload:
        $ref: '#/components/schemas/PostChapterizationMessage'
    postSummarization:
      name: post_summarization
      title: Summarization
      contentType: application/json
      summary: >-
        This message contains the informations about the post summarization of
        the audio chunk.
      payload:
        $ref: '#/components/schemas/PostSummarizationMessage'
    audioChunkAck:
      name: audio_chunk_ack
      title: Audio Chunk Acknowledgment
      contentType: application/json
      summary: >-
        This message contains the informations about the audio chunk
        acknowledgment of the audio chunk.
      payload:
        $ref: '#/components/schemas/AudioChunkAckMessage'
    stopRecordingAck:
      name: stop_recording_ack
      title: Stop Recording Acknowledgment
      contentType: application/json
      summary: >-
        This message contains the informations about the stop recording
        acknowledgment of the audio chunk.
      payload:
        $ref: '#/components/schemas/StopRecordingAckMessage'
    startSession:
      name: start_session
      title: Start Session
      contentType: application/json
      summary: >-
        This message contains the informations about the start session of the
        audio chunk.
      payload:
        $ref: '#/components/schemas/SimpleLifecycleMessage'
    startRecording:
      name: start_recording
      title: Start Recording
      contentType: application/json
      summary: >-
        This message contains the informations about the start recording of the
        audio chunk.
      payload:
        $ref: '#/components/schemas/SimpleLifecycleMessageStart'
    endRecording:
      name: end_recording
      title: End Recording
      contentType: application/json
      summary: >-
        This message contains the informations about the end recording of the
        audio chunk.
      payload:
        $ref: '#/components/schemas/EndRecordingMessage'
    endSession:
      name: end_session
      title: End Session
      contentType: application/json
      summary: >-
        This message contains the informations about the end session of the
        audio chunk.
      payload:
        $ref: '#/components/schemas/SimpleLifecycleMessageEndSession'
  schemas:
    UUID:
      type: string
      format: uuid
      description: UUID identifying the session.
      examples:
        - 550e8400-e29b-41d4-a716-446655440000
    ISODateTime:
      type: string
      format: date-time
      description: ISO 8601 timestamp in UTC.
      examples:
        - '2025-09-19T12:34:56Z'
    Word:
      type: object
      properties:
        word:
          type: string
          description: Recognized token as spoken in the audio.
        start:
          type: number
          description: Start time of the word in seconds from session start.
        end:
          type: number
          description: End time of the word in seconds from session start.
        confidence:
          type: number
          description: Confidence score for the token between 0 and 1.
      description: Word-level timing and confidence for a token in the transcript.
      examples:
        - word: hello
          start: 0.12
          end: 0.38
          confidence: 0.97
    Utterance:
      type: object
      properties:
        start:
          type: number
          description: Start time of the utterance in seconds from session start.
        end:
          type: number
          description: End time of the utterance in seconds from session start.
        confidence:
          type: number
          description: Confidence score for the utterance between 0 and 1.
        channel:
          type: integer
          description: Audio channel index (0-based) if multiple channels are present.
        words:
          type: array
          items:
            $ref: '#/components/schemas/Word'
          description: Word-level details for this utterance.
        text:
          type: string
          description: The transcribed text for the utterance.
        language:
          type: string
          description: BCP-47 language tag detected for the utterance (e.g. "en").
      description: Continuous segment of speech with timing, text and optional channel.
      examples:
        - start: 1.02
          end: 3.44
          confidence: 0.93
          channel: 0
          words:
            - word: good
              start: 1.02
              end: 1.21
              confidence: 0.96
            - word: morning
              start: 1.22
              end: 1.55
              confidence: 0.95
          text: good morning
          language: en
    ErrorObject:
      type: object
      properties:
        status_code:
          oneOf:
            - type: integer
            - type: string
          description: HTTP-like code or provider-specific status string.
        exception:
          type: string
          description: Short machine-readable error name.
        message:
          type: string
          description: Human-readable error explanation.
      description: Standard error container included in payloads when an operation fails.
      examples:
        - status_code: 400
          exception: InvalidArgument
          message: Unsupported sample rate
    TranscriptMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: transcript
        data:
          type: object
          required:
            - id
            - is_final
            - utterance
          properties:
            id:
              type: string
            is_final:
              type: boolean
              description: Indicates if the transcript is final or partial.
            utterance:
              $ref: '#/components/schemas/Utterance'
      description: Realtime transcript update for an utterance; may be interim or final.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:10Z'
          type: transcript
          data:
            id: 00-00000011
            is_final: true
            utterance:
              start: 0
              end: 0.48
              confidence: 0.91
              channel: 0
              words:
                - word: Hello
                  start: 0
                  end: 0.35
                  confidence: 0.91
                - word: world
                  start: 0.36
                  end: 0.48
                  confidence: 0.91
              text: Hello world.
              language: en
    SpeechEventData:
      type: object
      properties:
        time:
          type: number
          description: Time in seconds from session start when the event occurred.
        channel:
          type: integer
          description: Audio channel index (0-based) for which the event applies.
      description: Timing information for speech start/end events.
      examples:
        - time: 1.24
          channel: 0
    SpeechEventMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: speech_start
        data:
          $ref: '#/components/schemas/SpeechEventData'
      description: Indicates the start of detected speech activity on a given channel.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:06Z'
          type: speech_start
          data:
            time: 1.24
            channel: 0
    SpeechEventMessageEnd:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: speech_end
        data:
          $ref: '#/components/schemas/SpeechEventData'
      description: Indicates the end of detected speech activity.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:09Z'
          type: speech_end
          data:
            time: 3.1
            channel: 0
    SpeechEndMessage:
      title: Speech End
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: speech_end
        data:
          type: object
          properties:
            time:
              type: number
              description: Time in seconds from session start when the event occurred.
            channel:
              type: integer
              description: Audio channel index (0-based) for which the event applies.
          description: Timing information for the end of detected speech.
      description: >-
        Indicates the end of detected speech activity on a channel, with timing
        metadata.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:09Z'
          type: speech_end
          data:
            time: 3.1
            channel: 0
    SpeechStartMessage:
      title: Speech Start
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: speech_start
        data:
          type: object
          properties:
            time:
              type: number
              description: Time in seconds from session start when the event occurred.
            channel:
              type: integer
              description: Audio channel index (0-based) for which the event applies.
          description: Timing information for the start of detected speech.
      description: >-
        Indicates the start of detected speech activity on a channel, with
        timing metadata.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:06Z'
          type: speech_start
          data:
            time: 1.24
            channel: 0
    TranslationMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: translation
        data:
          type: object
          required:
            - utterance_id
            - utterance
            - original_language
            - target_language
            - translated_utterance
          properties:
            utterance_id:
              type: string
            utterance:
              $ref: '#/components/schemas/Utterance'
            original_language:
              type: string
            target_language:
              type: string
            translated_utterance:
              $ref: '#/components/schemas/Utterance'
      description: Translation result for an utterance into the requested target language.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:20Z'
          type: translation
          error: null
          data:
            utterance_id: utt_001
            utterance:
              text: buenos días
              language: es
              start: 4.2
              end: 6.1
            original_language: es
            target_language: en
            translated_utterance:
              text: good morning
              language: en
              start: 4.2
              end: 6.1
    NamedEntityRecognitionMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: named_entity_recognition
        data:
          type: object
          required:
            - utterance_id
            - utterance
            - results
          properties:
            utterance_id:
              type: string
            utterance:
              $ref: '#/components/schemas/Utterance'
            results:
              type: array
              items:
                type: object
                properties:
                  entity_type:
                    type: string
                  text:
                    type: string
                  start:
                    type: number
                  end:
                    type: number
      description: Named entity recognition results extracted from the given utterance.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:30Z'
          type: named_entity_recognition
          error: null
          data:
            utterance_id: utt_002
            utterance:
              text: meeting with Alice at 3pm
              start: 7
              end: 10.5
            results:
              - entity_type: PERSON
                text: Alice
                start: 2
                end: 2.5
              - entity_type: TIME
                text: 3pm
                start: 3
                end: 3.3
    SentimentAnalysisMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: sentiment_analysis
        data:
          type: object
          required:
            - utterance_id
            - utterance
            - results
          properties:
            utterance_id:
              type: string
            utterance:
              $ref: '#/components/schemas/Utterance'
            results:
              type: array
              items:
                type: object
                properties:
                  sentiment:
                    type: string
                  emotion:
                    type: string
                  text:
                    type: string
                  start:
                    type: number
                  end:
                    type: number
                  channel:
                    type: integer
      description: Sentiment and emotion analysis associated with segments of an utterance.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:35Z'
          type: sentiment_analysis
          error: null
          data:
            utterance_id: utt_003
            utterance:
              text: I love this product!
              start: 11
              end: 13.2
            results:
              - sentiment: positive
                emotion: joy
                text: love this product
                start: 11.4
                end: 12.9
                channel: 0
    SentenceResult:
      type: object
      properties:
        success:
          type: boolean
        is_empty:
          type: boolean
        exec_time:
          type: number
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        results:
          type: array
          items:
            type: string
      description: >-
        Result container for sentence-level operations such as punctuation or
        translation.
      examples:
        - success: true
          is_empty: false
          exec_time: 0.12
          error: null
          results:
            - Hello, world.
    SubtitleBlock:
      type: object
      properties:
        format:
          type: string
          enum:
            - srt
            - vtt
            - ass
        subtitles:
          type: string
      description: Subtitle payload for a specific format.
      examples:
        - format: vtt
          subtitles: |
            WEBVTT

            00:00:00.000 --> 00:00:01.000
            Hello world
    PostTranscriptData:
      type: object
      properties:
        full_transcript:
          type: string
        languages:
          type: array
          items:
            type: string
        sentences:
          type: array
          items:
            $ref: '#/components/schemas/SentenceResult'
        subtitles:
          type: array
          items:
            $ref: '#/components/schemas/SubtitleBlock'
        utterances:
          type: array
          items:
            $ref: '#/components/schemas/Utterance'
      description: Aggregated transcription artifacts produced after the session completes.
      examples:
        - full_transcript: hello world
          languages:
            - en
          sentences:
            - success: true
              is_empty: false
              exec_time: 0.11
              error: null
              results:
                - Hello world.
          subtitles:
            - format: srt
              subtitles: |
                1
                00:00:00,000 --> 00:00:01,000
                Hello world
          utterances:
            - start: 0
              end: 0.9
              text: hello world
    PostTranscriptMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: post_transcript
        data:
          $ref: '#/components/schemas/PostTranscriptData'
      description: Post-processing transcript payload containing full text and artifacts.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:40:00Z'
          type: post_transcript
          data:
            full_transcript: hello world
    PostFinalTranscriptMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: post_final_transcript
        data:
          type: object
          properties:
            metadata:
              type: object
              properties:
                audio_duration:
                  type: number
                number_of_distinct_channels:
                  type: integer
                billing_time:
                  type: number
                transcription_time:
                  type: number
            transcription:
              $ref: '#/components/schemas/PostTranscriptData'
            translation:
              type: object
              properties:
                success:
                  type: boolean
                is_empty:
                  type: boolean
                exec_time:
                  type: number
                error:
                  oneOf:
                    - type: 'null'
                    - $ref: '#/components/schemas/ErrorObject'
                results:
                  type: array
                  items:
                    type: object
                    properties:
                      error:
                        oneOf:
                          - type: 'null'
                          - $ref: '#/components/schemas/ErrorObject'
                      full_transcript:
                        type: string
                      languages:
                        type: array
                        items:
                          type: string
                      sentences:
                        type: array
                        items:
                          $ref: '#/components/schemas/SentenceResult'
                      subtitles:
                        type: array
                        items:
                          $ref: '#/components/schemas/SubtitleBlock'
                      utterances:
                        type: array
                        items:
                          $ref: '#/components/schemas/Utterance'
            summarization:
              type: object
              properties:
                success:
                  type: boolean
                is_empty:
                  type: boolean
                exec_time:
                  type: number
                error:
                  oneOf:
                    - type: 'null'
                    - $ref: '#/components/schemas/ErrorObject'
                results:
                  type: string
            named_entity_recognition:
              type: object
              properties:
                success:
                  type: boolean
                is_empty:
                  type: boolean
                exec_time:
                  type: number
                error:
                  oneOf:
                    - type: 'null'
                    - $ref: '#/components/schemas/ErrorObject'
                entity:
                  type: string
            sentiment_analysis:
              type: object
              properties:
                success:
                  type: boolean
                is_empty:
                  type: boolean
                exec_time:
                  type: number
                error:
                  oneOf:
                    - type: 'null'
                    - $ref: '#/components/schemas/ErrorObject'
                results:
                  type: string
            chapterization:
              type: object
              properties:
                success:
                  type: boolean
                is_empty:
                  type: boolean
                exec_time:
                  type: number
                error:
                  oneOf:
                    - type: 'null'
                    - $ref: '#/components/schemas/ErrorObject'
                results:
                  type: object
      description: >-
        Final aggregated output for the session, including metadata and add-on
        results.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:45:00Z'
          type: post_final_transcript
          data:
            metadata:
              audio_duration: 123.45
              number_of_distinct_channels: 1
              billing_time: 120
              transcription_time: 3.2
            transcription:
              full_transcript: hello world
    PostChapterizationMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: post_chapterization
        data:
          type: object
          properties:
            results:
              type: array
              items:
                type: object
                properties:
                  abstractive_summary:
                    type: string
                  extractive_summary:
                    type: string
                  summary:
                    type: string
                  headline:
                    type: string
                  gist:
                    type: string
                  keywords:
                    type: array
                    items:
                      type: string
                  start:
                    type: number
                  end:
                    type: number
                  sentences:
                    type: array
                    items:
                      type: object
                      properties:
                        sentence:
                          type: string
                        start:
                          type: number
                        end:
                          type: number
                        words:
                          type: array
                          items:
                            $ref: '#/components/schemas/Word'
                  text:
                    type: string
      description: Post-processing output with chapter-level summaries and metadata.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:46:00Z'
          type: post_chapterization
          error: null
          data:
            results:
              - headline: Project kickoff
                start: 0
                end: 60
                keywords:
                  - timeline
                  - owner
    PostSummarizationMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: post_summarization
        data:
          type: object
          properties:
            results:
              type: string
      description: Post-processing abstractive summary of the conversation.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:47:00Z'
          type: post_summarization
          error: null
          data:
            results: The team aligned on goals and next steps for the quarter.
    AudioChunkAckMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - acknowledged
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        acknowledged:
          type: boolean
          description: >-
            True if the server successfully received and buffered the audio
            chunk.
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: audio_chunk
          description: Event type for audio chunk acknowledgments.
        data:
          type: object
          properties:
            byte_range:
              type: array
              items:
                type: number
              minItems: 2
              maxItems: 2
              description: Inclusive byte offsets [start, end] acknowledged for this chunk.
            time_range:
              type: array
              items:
                type: number
              minItems: 2
              maxItems: 2
              description: >-
                Inclusive audio time range in seconds [start, end] covered by
                this chunk.
      description: >-
        Acknowledges receipt of an audio chunk and indicates the acknowledged
        byte/time ranges.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:56Z'
          type: audio_chunk
          acknowledged: true
          error: null
          data:
            byte_range:
              - 0
              - 4095
            time_range:
              - 0
              - 0.256
    StopRecordingAckMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - acknowledged
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        acknowledged:
          type: boolean
          description: True if stop_recording was accepted; false if rejected.
        error:
          oneOf:
            - type: 'null'
            - $ref: '#/components/schemas/ErrorObject'
        type:
          type: string
          const: stop_recording
          description: Event type for stop recording acknowledgments.
        data:
          type: object
          description: >-
            Additional context provided by the provider; may be empty when
            acknowledged is true.
      description: Acknowledges the stop recording command with optional provider details.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:35:10Z'
          type: stop_recording
          acknowledged: true
          error: null
          data: {}
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:35:12Z'
          type: stop_recording
          acknowledged: false
          error:
            status_code: 409
            exception: ConflictError
            message: Recording already stopped
          data: {}
    SimpleLifecycleMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: start_session
          description: Lifecycle event emitted once when the session is created and ready.
      description: Base lifecycle event for session start.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:00Z'
          type: start_session
    SimpleLifecycleMessageStart:
      type: object
      required:
        - session_id
        - created_at
        - type
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: start_recording
          description: Lifecycle event emitted when audio capture/streaming begins.
      description: Lifecycle event indicating the start of recording.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:34:05Z'
          type: start_recording
    EndRecordingMessage:
      type: object
      required:
        - session_id
        - created_at
        - type
        - data
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: end_recording
          description: >-
            Lifecycle event emitted when the provider has stopped receiving
            audio.
        data:
          type: object
          additionalProperties: true
          description: >-
            Provider-defined object with optional metadata related to the end of
            recording.
      description: >-
        Lifecycle event sent when recording has ended, optionally carrying
        provider metadata.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:36:00Z'
          type: end_recording
          data:
            reason: user_request
            received_total_bytes: 1048576
    SimpleLifecycleMessageEndSession:
      type: object
      required:
        - session_id
        - created_at
        - type
      properties:
        session_id:
          $ref: '#/components/schemas/UUID'
        created_at:
          $ref: '#/components/schemas/ISODateTime'
        type:
          type: string
          const: end_session
          description: >-
            Lifecycle event emitted when the session is closed and no further
            data will be sent.
      description: Lifecycle event for session termination.
      examples:
        - session_id: 550e8400-e29b-41d4-a716-446655440000
          created_at: '2025-09-19T12:36:05Z'
          type: end_session