Hyperbolic Audio Generation API

openapi: 3.1.0
info:
  title: Hyperbolic Audio Generation API
  description: >
    Convert text to natural-sounding speech using audio models hosted by
    Hyperbolic — Melo TTS (sunset) and Whisper (coming soon). Returns
    base64-encoded audio. Pricing from $0.001 per 1000 characters.
  version: v1
  contact:
    name: Hyperbolic Support
    email: support@hyperbolic.ai
    url: https://docs.hyperbolic.ai
  license:
    name: Hyperbolic Terms of Use
    url: https://www.hyperbolic.ai/terms-of-use

servers:
  - url: https://api.hyperbolic.xyz/v1
    description: Hyperbolic Production Inference Server

security:
  - BearerAuth: []

tags:
  - name: Audio Generation
    description: Text-to-speech audio endpoint

paths:
  /audio/generation:
    post:
      summary: Hyperbolic Generate Audio
      description: >
        Generate speech audio from text using the selected TTS model.
      operationId: generateAudio
      tags:
        - Audio Generation
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/AudioGenerationRequest'
            examples:
              BasicTTS:
                summary: Basic TTS request
                value:
                  text: Hello and welcome to Hyperbolic.
                  speed: 1.0
                  language: EN
      responses:
        '200':
          description: Successful audio generation
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AudioGenerationResponse'
        '400':
          description: Bad Request
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Unauthorized
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Too Many Requests
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'

components:
  securitySchemes:
    BearerAuth:
      type: http
      scheme: bearer
      bearerFormat: API Key

  schemas:
    AudioGenerationRequest:
      type: object
      required:
        - text
      properties:
        text:
          type: string
          description: Input text to convert to speech.
        speed:
          type: number
          minimum: 0.5
          maximum: 2.0
          default: 1.0
        language:
          type: string
          description: Language code (e.g. EN, ES, FR, JA, ZH, KR).
          default: EN
        voice:
          type: string
          description: Speaker / voice ID for the selected model.

    AudioGenerationResponse:
      type: object
      properties:
        audio:
          type: string
          description: Base64-encoded audio bytes (typically WAV or MP3).
        duration:
          type: number
          description: Audio length in seconds.

    ErrorResponse:
      type: object
      properties:
        error:
          type: object
          properties:
            message:
              type: string
            type:
              type: string
            code:
              type: string