Camb.AI Voices API

openapi: 3.0.3
info:
  title: Camb.AI API
  description: >-
    The Camb.AI API provides generative voice AI - text-to-speech, dubbing,
    translation, voice discovery and cloning, and transcription - powered by the
    MARS (speech / voice) and BOLI (translation) research models. The REST API is
    served from https://client.camb.ai/apis and authenticated with an x-api-key
    header. Most operations are asynchronous: a POST starts a task and returns a
    task_id, the task is polled with a GET until its status is SUCCESS, and the
    result (audio, transcript, or translation) is then retrieved. Real-time
    streaming for TTS, transcription, and speech-to-speech translation is offered
    over separate public WebSocket channels (see the companion AsyncAPI document).
    Endpoint paths are grounded in the public documentation at https://docs.camb.ai;
    request and response schemas below are modeled from the documented parameters
    and may be simplified.
  version: '1.0'
  contact:
    name: Camb.AI
    url: https://www.camb.ai
servers:
  - url: https://client.camb.ai/apis
    description: Camb.AI production API
security:
  - apiKeyAuth: []
tags:
  - name: Text-to-Speech
    description: Convert text to speech with the MARS voice models.
  - name: Dubbing
    description: End-to-end video and audio dubbing across languages.
  - name: Translation
    description: Neural text translation via the BOLI model.
  - name: Voices
    description: Discover, clone, design, and manage voices.
  - name: Transcription
    description: Speech-to-text transcription with timestamps.
  - name: Languages
    description: Supported source and target languages.
paths:
  /tts-stream:
    post:
      operationId: createTtsStream
      tags:
        - Text-to-Speech
      summary: Stream text-to-speech
      description: >-
        Synthesize speech from text and return a binary audio stream in real time.
        The response is an audio byte stream (for example audio/wav or audio/mpeg)
        and includes an X-Credits-Required header indicating credit usage.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TtsStreamRequest'
      responses:
        '200':
          description: A binary audio stream.
          content:
            audio/wav:
              schema:
                type: string
                format: binary
        '401':
          $ref: '#/components/responses/Unauthorized'
  /tts:
    post:
      operationId: createTts
      deprecated: true
      tags:
        - Text-to-Speech
      summary: Create a text-to-speech task (deprecated)
      description: >-
        Deprecated. For new integrations use POST /tts-stream. Starts an
        asynchronous text-to-speech task and returns a task_id to poll.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TtsRequest'
      responses:
        '200':
          description: Task accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /tts/{task_id}:
    get:
      operationId: getTtsStatus
      tags:
        - Text-to-Speech
      summary: Poll a text-to-speech task
      description: Returns the status of a TTS task and, when complete, the run_id used to fetch the audio.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Current task status.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskStatus'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /tts-result/{run_id}:
    get:
      operationId: getTtsResult
      tags:
        - Text-to-Speech
      summary: Download text-to-speech audio
      description: Retrieves the generated audio file for a completed TTS run.
      parameters:
        - name: run_id
          in: path
          required: true
          description: The run identifier returned by GET /tts/{task_id}.
          schema:
            type: integer
      responses:
        '200':
          description: The generated audio file.
          content:
            audio/wav:
              schema:
                type: string
                format: binary
        '401':
          $ref: '#/components/responses/Unauthorized'
  /dub:
    post:
      operationId: createDub
      tags:
        - Dubbing
      summary: Create an end-to-end dubbing task
      description: >-
        Starts an end-to-end dubbing task for a media file, translating and
        re-voicing it into one or more target languages while preserving delivery.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/DubRequest'
      responses:
        '200':
          description: Dubbing task accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /dub/{task_id}:
    get:
      operationId: getDubStatus
      tags:
        - Dubbing
      summary: Poll a dubbing task
      description: Returns the status and, when complete, the results of a dubbing task.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Current dubbing task status and results.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskStatus'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /translate:
    post:
      operationId: createTranslation
      tags:
        - Translation
      summary: Create a translation task
      description: >-
        Translates an array of texts from a source language into a target
        language using the BOLI model, with optional formality, gender, and
        specialized-vocabulary controls.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TranslationRequest'
      responses:
        '200':
          description: Translation task accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /translate/{task_id}:
    get:
      operationId: getTranslationStatus
      tags:
        - Translation
      summary: Poll a translation task
      description: Returns the status and, when complete, the translated texts.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Current translation task status and result.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskStatus'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /transcribe:
    post:
      operationId: createTranscription
      tags:
        - Transcription
      summary: Create a transcription task
      description: >-
        Transcribes speech to text from an uploaded media file or a media URL,
        with speaker identification and word-level timestamps.
      requestBody:
        required: true
        content:
          multipart/form-data:
            schema:
              $ref: '#/components/schemas/TranscriptionRequest'
      responses:
        '200':
          description: Transcription task accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /transcribe/{task_id}:
    get:
      operationId: getTranscriptionStatus
      tags:
        - Transcription
      summary: Poll a transcription task
      description: Returns the status and, when complete, the transcript.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Current transcription task status and result.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskStatus'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /list-voices:
    get:
      operationId: listVoices
      tags:
        - Voices
      summary: List voices
      description: >-
        Returns all available voices - public voices, shared voices, and
        custom-created voices - with metadata such as id, voice_name, gender,
        age, language, and description.
      responses:
        '200':
          description: A list of voices.
          content:
            application/json:
              schema:
                type: array
                items:
                  $ref: '#/components/schemas/Voice'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /create-custom-voice:
    post:
      operationId: createCustomVoice
      tags:
        - Voices
      summary: Create a custom (cloned) voice
      description: Clones a custom voice from a reference audio sample.
      requestBody:
        required: true
        content:
          multipart/form-data:
            schema:
              $ref: '#/components/schemas/CustomVoiceRequest'
      responses:
        '200':
          description: The created custom voice.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Voice'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /text-to-voice:
    post:
      operationId: createTextToVoice
      tags:
        - Voices
      summary: Design a voice from a text description
      description: Starts a task that generates candidate voices from a natural-language description.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TextToVoiceRequest'
      responses:
        '200':
          description: Voice-design task accepted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskCreated'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /text-to-voice/{task_id}:
    get:
      operationId: getTextToVoiceStatus
      tags:
        - Voices
      summary: Poll a voice-design task
      description: Returns the status and, when complete, the generated voice options.
      parameters:
        - $ref: '#/components/parameters/TaskId'
      responses:
        '200':
          description: Current voice-design task status and result.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TaskStatus'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /delete-voice/{voice_id}:
    delete:
      operationId: deleteVoice
      tags:
        - Voices
      summary: Delete a custom voice
      description: Permanently removes a custom voice.
      parameters:
        - name: voice_id
          in: path
          required: true
          description: The identifier of the custom voice to delete.
          schema:
            type: integer
      responses:
        '200':
          description: Voice deleted.
        '401':
          $ref: '#/components/responses/Unauthorized'
  /source-languages:
    get:
      operationId: getSourceLanguages
      tags:
        - Languages
      summary: List source languages
      description: Lists all supported source languages with identifiers and locale tags.
      responses:
        '200':
          description: A list of source languages.
          content:
            application/json:
              schema:
                type: array
                items:
                  $ref: '#/components/schemas/Language'
        '401':
          $ref: '#/components/responses/Unauthorized'
  /target-languages:
    get:
      operationId: getTargetLanguages
      tags:
        - Languages
      summary: List target languages
      description: Lists all supported target languages with identifiers and locale tags.
      responses:
        '200':
          description: A list of target languages.
          content:
            application/json:
              schema:
                type: array
                items:
                  $ref: '#/components/schemas/Language'
        '401':
          $ref: '#/components/responses/Unauthorized'
components:
  securitySchemes:
    apiKeyAuth:
      type: apiKey
      in: header
      name: x-api-key
  parameters:
    TaskId:
      name: task_id
      in: path
      required: true
      description: The task identifier returned when the task was created.
      schema:
        type: string
  responses:
    Unauthorized:
      description: Missing or invalid API key.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
  schemas:
    TtsStreamRequest:
      type: object
      required:
        - text
        - language
        - voice_id
      properties:
        text:
          type: string
          minLength: 3
          maxLength: 3000
          description: The text to synthesize into speech.
        language:
          type: string
          description: BCP-47 locale code, for example en-us or fr-fr.
        voice_id:
          type: integer
          description: A voice profile ID from GET /list-voices.
        speech_model:
          type: string
          description: Model variant, for example mars-8.1-flash-beta or mars-instruct.
        user_instructions:
          type: string
          description: Guidance for overall delivery (mars-instruct only).
        enhance_named_entities_pronunciation:
          type: boolean
        output_configuration:
          type: object
          description: Output format, sample rate, and enhancement settings.
        voice_settings:
          type: object
          description: Speaking rate, accent preservation, and reference quality.
        inference_options:
          type: object
          description: Stability, temperature, and speaker similarity.
    TtsRequest:
      type: object
      required:
        - text
        - voice_id
        - language
      properties:
        text:
          type: string
          description: The text to be converted to speech.
        voice_id:
          type: integer
          description: Voice identifier.
        language:
          type: string
          description: Locale tag such as en-us.
        gender:
          type: integer
          description: Speaker gender hint (0, 1, 2, or 9).
        age:
          type: integer
          description: Speaker age characteristic.
    DubRequest:
      type: object
      required:
        - video_url
        - source_language
        - target_languages
      properties:
        video_url:
          type: string
          description: URL of the media file to dub.
        source_language:
          type: string
          description: Source language locale tag, for example en-us.
        target_languages:
          type: array
          description: Target language locale tags, for example ["es-es", "fr-fr"].
          items:
            type: string
    TranslationRequest:
      type: object
      required:
        - texts
        - source_language
        - target_language
      properties:
        texts:
          type: array
          description: An array of strings to be translated.
          items:
            type: string
        source_language:
          type: string
          description: Source language locale tag.
        target_language:
          type: string
          description: Target language locale tag.
        formality:
          type: string
          description: Optional formality control.
        gender:
          type: string
          description: Optional gender control.
    TranscriptionRequest:
      type: object
      required:
        - language
      properties:
        language:
          type: string
          description: Locale tag of the audio, for example en-us.
        media_file:
          type: string
          format: binary
          description: Binary audio or video file (MP3, WAV, AAC, FLAC, MP4, MOV, MXF).
        media_url:
          type: string
          description: Media URL to transcribe (alternative to media_file).
        project_name:
          type: string
        project_description:
          type: string
    CustomVoiceRequest:
      type: object
      required:
        - voice_name
        - file
      properties:
        voice_name:
          type: string
          description: Name for the new custom voice.
        file:
          type: string
          format: binary
          description: Reference audio sample to clone from.
        gender:
          type: integer
        age:
          type: integer
    TextToVoiceRequest:
      type: object
      required:
        - voice_description
      properties:
        voice_description:
          type: string
          description: Natural-language description of the desired voice.
        text:
          type: string
          description: Optional preview text to render with the generated voices.
    TaskCreated:
      type: object
      properties:
        task_id:
          type: string
          description: Identifier used to poll the task.
    TaskStatus:
      type: object
      properties:
        status:
          type: string
          description: Task status, for example PENDING, PROCESSING, SUCCESS, or FAILED.
          enum:
            - PENDING
            - PROCESSING
            - SUCCESS
            - FAILED
        run_id:
          type: integer
          description: Present when the task has completed; used to fetch the result.
    Voice:
      type: object
      properties:
        id:
          type: integer
        voice_name:
          type: string
        gender:
          type: integer
        age:
          type: integer
        language:
          type: string
        description:
          type: string
        is_published:
          type: boolean
    Language:
      type: object
      properties:
        id:
          type: integer
        language:
          type: string
        short_name:
          type: string
    Error:
      type: object
      properties:
        message:
          type: string
        code:
          type: string