elevenlabs Speech to Speech API

openapi: 3.1.0
info:
  title: ElevenLabs Audio Isolation Agents Speech to Speech API
  description: The ElevenLabs Audio Isolation API removes background noise from audio recordings, isolating vocal tracks from ambient sounds and interference. This is useful for cleaning up recordings, improving audio quality for podcasts and interviews, and preparing audio files for further processing such as voice cloning or transcription.
  version: '1.0'
  contact:
    name: ElevenLabs Support
    url: https://help.elevenlabs.io
  termsOfService: https://elevenlabs.io/terms-of-service
servers:
- url: https://api.elevenlabs.io
  description: Production Server
security:
- apiKeyAuth: []
tags:
- name: Speech to Speech
  description: Endpoints for converting speech from one voice to another while preserving the original speech characteristics.
paths:
  /v1/speech-to-speech/{voice_id}:
    post:
      operationId: convertVoice
      summary: Voice changer
      description: Converts an audio recording to use a different voice while preserving the original speech content, timing, and emotional delivery. The target voice can be any voice available in the user's library.
      tags:
      - Speech to Speech
      parameters:
      - $ref: '#/components/parameters/voiceId'
      - $ref: '#/components/parameters/outputFormat'
      requestBody:
        required: true
        content:
          multipart/form-data:
            schema:
              $ref: '#/components/schemas/SpeechToSpeechRequest'
      responses:
        '200':
          description: Voice conversion completed successfully
          content:
            audio/mpeg:
              schema:
                type: string
                format: binary
        '400':
          description: Bad request - invalid audio or parameters
        '401':
          description: Unauthorized - invalid or missing API key
        '422':
          description: Unprocessable entity - audio could not be processed
  /v1/speech-to-speech/{voice_id}/stream:
    post:
      operationId: streamConvertedVoice
      summary: Voice changer stream
      description: Converts an audio recording to use a different voice and streams the result using chunked transfer encoding. Useful for real-time processing and playback scenarios.
      tags:
      - Speech to Speech
      parameters:
      - $ref: '#/components/parameters/voiceId'
      - $ref: '#/components/parameters/outputFormat'
      requestBody:
        required: true
        content:
          multipart/form-data:
            schema:
              $ref: '#/components/schemas/SpeechToSpeechRequest'
      responses:
        '200':
          description: Streaming voice conversion response
          content:
            audio/mpeg:
              schema:
                type: string
                format: binary
        '400':
          description: Bad request - invalid audio or parameters
        '401':
          description: Unauthorized - invalid or missing API key
        '422':
          description: Unprocessable entity - audio could not be processed
components:
  schemas:
    SpeechToSpeechRequest:
      type: object
      required:
      - audio
      properties:
        audio:
          type: string
          format: binary
          description: The source audio file containing the speech to convert. Supports common audio formats including MP3, WAV, and OGG.
        model_id:
          type: string
          description: The identifier of the model to use for voice conversion.
        voice_settings:
          type: object
          description: Voice settings to override the default settings for the target voice.
          properties:
            stability:
              type: number
              description: Controls the stability of the converted voice output.
              minimum: 0
              maximum: 1
            similarity_boost:
              type: number
              description: Controls how closely the output matches the target voice.
              minimum: 0
              maximum: 1
            style:
              type: number
              description: Controls the expressiveness of the converted speech.
              minimum: 0
              maximum: 1
            use_speaker_boost:
              type: boolean
              description: Enables speaker boost for increased clarity.
        seed:
          type: integer
          description: A seed value for deterministic generation.
  parameters:
    voiceId:
      name: voice_id
      in: path
      required: true
      description: The identifier of the target voice to convert the audio to.
      schema:
        type: string
    outputFormat:
      name: output_format
      in: query
      required: false
      description: The desired output audio format.
      schema:
        type: string
        default: mp3_44100_128
        enum:
        - mp3_22050_32
        - mp3_44100_32
        - mp3_44100_64
        - mp3_44100_96
        - mp3_44100_128
        - mp3_44100_192
        - pcm_16000
        - pcm_22050
        - pcm_24000
        - pcm_44100
        - ulaw_8000
  securitySchemes:
    apiKeyAuth:
      type: apiKey
      in: header
      name: xi-api-key
      description: ElevenLabs API key passed in the xi-api-key header for authentication.
externalDocs:
  description: ElevenLabs Audio Isolation API Documentation
  url: https://elevenlabs.io/docs/api-reference/audio-isolation/audio-isolation