Cartesia Speech-to-Text API

Batch transcription of an audio file of any length.

OpenAPI Specification

cartesia-ai-speech-to-text-api-openapi.yml Raw ↑
openapi: 3.0.3
info:
  title: Cartesia Agents Speech-to-Text API
  description: 'The Cartesia REST API for real-time voice AI - text-to-speech and voice-changer generation (single-shot bytes or Server-Sent Events), batch speech-to-text transcription, voice management and cloning, audio infill, custom datasets, pronunciation dictionaries, API keys and scoped access tokens, the Voice Agents platform (agents, calls, call batches, phone numbers, telephony providers, webhooks, deployments, knowledge base documents/folders, and evaluation metrics), and usage reporting. All requests require a `Cartesia-Version` header and an `Authorization: Bearer` API key (`sk_car_...`) or short-lived access token. The lowest-latency, streaming surface for TTS and STT is a separate WebSocket protocol documented in the companion AsyncAPI document at asyncapi/cartesia-ai-asyncapi.yml.'
  version: '2026-03-01'
  contact:
    name: Cartesia
    url: https://cartesia.ai
  license:
    name: API documentation - Cartesia Terms of Service
    url: https://cartesia.ai/terms-of-service
servers:
- url: https://api.cartesia.ai
  description: Cartesia production API
security:
- bearerAuth: []
tags:
- name: Speech-to-Text
  description: Batch transcription of an audio file of any length.
paths:
  /stt:
    post:
      operationId: transcribe
      tags:
      - Speech-to-Text
      summary: Batch Speech-to-Text
      description: Transcribes an audio file of any length using the ink-whisper model.
      parameters:
      - $ref: '#/components/parameters/CartesiaVersion'
      - name: encoding
        in: query
        required: false
        description: PCM encoding, required only when uploading raw headerless PCM.
        schema:
          type: string
          enum:
          - pcm_s16le
          - pcm_s32le
          - pcm_f16le
          - pcm_f32le
          - pcm_mulaw
          - pcm_alaw
      - name: sample_rate
        in: query
        required: false
        description: Sample rate in Hz, required only when uploading raw headerless PCM.
        schema:
          type: integer
      requestBody:
        required: true
        content:
          multipart/form-data:
            schema:
              type: object
              required:
              - file
              - model
              properties:
                file:
                  type: string
                  format: binary
                  description: 'Audio file. Supported formats: flac, m4a, mp3, mp4, mpeg.'
                model:
                  type: string
                  enum:
                  - ink-whisper
                language:
                  type: string
                  description: ISO-639-1 code. Defaults to en. Supports 99+ languages.
                  default: en
                timestamp_granularities[]:
                  type: array
                  items:
                    type: string
                    enum:
                    - word
      responses:
        '200':
          description: The transcript.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TranscriptResponse'
        '401':
          $ref: '#/components/responses/Unauthorized'
        '422':
          $ref: '#/components/responses/ValidationError'
components:
  responses:
    Unauthorized:
      description: Missing, invalid, or expired API key / access token.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
    ValidationError:
      description: The request payload failed validation.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
  schemas:
    TranscriptResponse:
      type: object
      required:
      - type
      - text
      properties:
        type:
          type: string
          enum:
          - transcript
        request_id:
          type: string
        text:
          type: string
        language:
          type: string
        duration:
          type: number
        words:
          type: array
          items:
            $ref: '#/components/schemas/WordTiming'
    WordTiming:
      type: object
      properties:
        word:
          type: string
        start:
          type: number
        end:
          type: number
    Error:
      type: object
      properties:
        title:
          type: string
        message:
          type: string
        error_code:
          type: string
        status_code:
          type: integer
        doc_url:
          type: string
        request_id:
          type: string
  parameters:
    CartesiaVersion:
      name: Cartesia-Version
      in: header
      required: true
      description: API version date, e.g. 2026-03-01.
      schema:
        type: string
        example: '2026-03-01'
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      bearerFormat: sk_car_... API key or short-lived access token
      description: 'Cartesia API key (sk_car_...) or a short-lived access token minted via POST /access-token, passed as Authorization: Bearer <token>. Every request also requires the Cartesia-Version header.'
Where this information came from

This is an independent, third-party profile of Cartesia Speech-to-Text API, published by API Evangelist. We do not operate, host, resell, or support these APIs, and we are not affiliated with or endorsed by the company unless stated above. Everything here is built from publicly available information — the company's own site, developer portal, documentation, public repositories, and the specifications it publishes for public use. Nothing is obtained by breaching a system, defeating an access control, or using credentials.

The Kin Score and Agent Readiness rating are independently calculated assessments of a company's public API artifacts, scored against a published rubric. They are not certifications, endorsements, security assessments, or audits.

Corrections, re-scores, and removal are free — no partnership or purchase required, and you do not need to justify the request. A removed company is recorded as unrated, never scored zero for having asked. Acknowledgement within one business day; removal within two.

info@apievangelist.com · Read the full data-sourcing policy →
On a security or compliance team? Put security in the subject line and you will get a person, not a form — we will tell you exactly which public URLs this profile was built from.