WellSaid Labs Text-to-Speech API

openapi: 3.0.3
info:
  title: WellSaid Labs Clips Text-to-Speech API
  description: The WellSaid Labs API renders text into studio-quality AI voiceover. It exposes streaming text-to-speech (audio returned as an HTTP stream for low time-to-first-byte playback), word-level timing and subtitle rendering, asynchronous clip creation and combination, a catalog of voice avatars, and pronunciation control via replacement libraries and respelling suggestions. All requests authenticate with an X-Api-Key header. WellSaid does not currently support end-user authentication and recommends making requests from an internal or trusted source. API access is gated behind a trial API key and a business plan. Endpoints, request fields, and response shapes here are modeled from the public documentation at docs.wellsaidlabs.com and are honestly approximated where the reference does not publish a full schema.
  version: '1.0'
  contact:
    name: WellSaid Labs
    url: https://wellsaidlabs.com
servers:
- url: https://api.wellsaidlabs.com/v1
  description: WellSaid Labs API
security:
- apiKeyAuth: []
tags:
- name: Text-to-Speech
  description: Render text to speech, with streaming audio and word timing.
paths:
  /tts/stream:
    post:
      operationId: renderTtsStream
      tags:
      - Text-to-Speech
      summary: Render text to speech (streaming)
      description: Renders the supplied text with the chosen voice avatar and returns the result as a stream of an audio file (MP3), not the finished file, so the client can begin playback with a low time-to-first-byte. Default keys are limited to 1,000 characters per request. Accepts a subset of SSML tags.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TtsRequest'
      responses:
        '200':
          description: A stream of the rendered audio.
          content:
            audio/mpeg:
              schema:
                type: string
                format: binary
        '401':
          $ref: '#/components/responses/Unauthorized'
        '429':
          $ref: '#/components/responses/RateLimited'
  /word-timing:
    post:
      operationId: renderWordTiming
      tags:
      - Text-to-Speech
      summary: Render text to speech with timing information
      description: Renders text to speech and returns the audio together with word-level timing information and subtitles, so captions can be aligned to the generated voiceover.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TtsRequest'
      responses:
        '200':
          description: Rendered audio with word timing and subtitle data.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/WordTimingResponse'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '429':
          $ref: '#/components/responses/RateLimited'
components:
  responses:
    RateLimited:
      description: Rate limit or monthly character quota exceeded. Inspect the x-rate-limit-* and x-quota-* response headers.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
    Unauthorized:
      description: Missing or invalid X-Api-Key.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
  schemas:
    Error:
      type: object
      properties:
        message:
          type: string
        code:
          type: string
    WordTimingResponse:
      type: object
      properties:
        audio_url:
          type: string
          description: URL to the rendered audio.
        subtitles:
          type: string
          description: Subtitle text aligned to the audio.
        words:
          type: array
          description: Per-word timing entries.
          items:
            type: object
            properties:
              word:
                type: string
              start:
                type: number
                format: float
                description: Start time in seconds.
              end:
                type: number
                format: float
                description: End time in seconds.
    TtsRequest:
      type: object
      properties:
        text:
          type: string
          description: The text to convert to speech. Default keys allow up to 1,000 characters per request.
        speaker_id:
          type: string
          description: The voice avatar (speaker) id to render with, for example "3".
        model:
          type: string
          description: The voice model to use, for example "caruso" or "legacy".
      required:
      - text
      - speaker_id
  securitySchemes:
    apiKeyAuth:
      type: apiKey
      in: header
      name: X-Api-Key
      description: API key issued by WellSaid Labs, passed in the X-Api-Key header. Request a trial key from the console; production use requires a business plan.