Streams synthesized audio as a chunked HTTP response from POST /v1/audio/stream for sub-300ms time-to-first-byte playback, with Accept values of audio/mpeg, audio/ogg, audio/aac, or audio/pcm (WAV is not available for streaming).
openapi: 3.0.1
info:
title: Speechify Voice API
description: >-
Specification of the Speechify Voice API (Text-to-Speech). The API
synthesizes lifelike speech from plain text or SSML in 30+ languages,
supports non-streaming and streaming audio, lists available voices, and
creates and deletes instant voice clones. All requests authenticate with a
Bearer API key.
termsOfService: https://speechify.com/terms/
contact:
name: Speechify Support
url: https://docs.sws.speechify.com
version: '1.0'
servers:
- url: https://api.sws.speechify.com
security:
- bearerAuth: []
paths:
/v1/audio/speech:
post:
operationId: createSpeech
tags:
- Audio
summary: Convert text to speech
description: >-
Synthesizes speech for the given input and returns a JSON object with
base64-encoded audio data, the audio format, the billable character
count, and speech marks describing word-level timing.
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/GetSpeechRequest'
responses:
'200':
description: OK
content:
application/json:
schema:
$ref: '#/components/schemas/GetSpeechResponse'
'401':
description: Unauthorized - missing or invalid API key.
'422':
description: Unprocessable entity - invalid request parameters.
/v1/audio/stream:
post:
operationId: streamSpeech
tags:
- Audio
summary: Stream text to speech
description: >-
Synthesizes speech for the given input and returns the raw audio as a
chunked HTTP stream for low-latency playback. The response content type
is selected with the Accept header. WAV is not available for streaming;
use /v1/audio/speech for WAV output.
parameters:
- name: Accept
in: header
required: false
description: >-
Desired audio container for the streamed response. Allowed values
are audio/mpeg, audio/ogg, audio/aac, or audio/pcm.
schema:
type: string
enum:
- audio/mpeg
- audio/ogg
- audio/aac
- audio/pcm
default: audio/mpeg
requestBody:
required: true
content:
application/json:
schema:
$ref: '#/components/schemas/GetStreamRequest'
responses:
'200':
description: A chunked stream of synthesized audio bytes.
headers:
Transfer-Encoding:
schema:
type: string
description: chunked
content:
audio/mpeg:
schema:
type: string
format: binary
audio/ogg:
schema:
type: string
format: binary
audio/aac:
schema:
type: string
format: binary
audio/pcm:
schema:
type: string
format: binary
'401':
description: Unauthorized - missing or invalid API key.
'422':
description: Unprocessable entity - invalid request parameters.
/v1/voices:
get:
operationId: listVoices
tags:
- Voices
summary: List voices
description: >-
Returns the available voices, including pre-set Speechify voices and
any personal cloned voices on the account.
responses:
'200':
description: OK
content:
application/json:
schema:
type: array
items:
$ref: '#/components/schemas/Voice'
'401':
description: Unauthorized - missing or invalid API key.
post:
operationId: createVoice
tags:
- Voices
summary: Create a voice clone
description: >-
Creates an instant personal voice clone from a 10-30 second audio
sample. Requires explicit consent identifying the person who owns the
voice. Submitted as multipart/form-data.
requestBody:
required: true
content:
multipart/form-data:
schema:
$ref: '#/components/schemas/CreateVoiceRequest'
responses:
'200':
description: The created personal voice.
content:
application/json:
schema:
$ref: '#/components/schemas/Voice'
'401':
description: Unauthorized - missing or invalid API key.
'422':
description: Unprocessable entity - invalid sample or missing consent.
/v1/voices/{id}:
delete:
operationId: deleteVoice
tags:
- Voices
summary: Delete a voice
description: Deletes a personal (cloned) voice by its id.
parameters:
- name: id
in: path
required: true
description: The id of the personal voice to delete.
schema:
type: string
responses:
'204':
description: The voice was deleted.
'401':
description: Unauthorized - missing or invalid API key.
'404':
description: The voice was not found.
components:
securitySchemes:
bearerAuth:
type: http
scheme: bearer
description: >-
Pass your Speechify API key as a Bearer token in the Authorization
header, e.g. "Authorization: Bearer $SPEECHIFY_API_KEY".
schemas:
GetSpeechRequest:
type: object
required:
- input
- voice_id
properties:
input:
type: string
description: Plain text or SSML to be synthesized to speech.
voice_id:
type: string
description: Id of the voice to be used for synthesizing speech.
language:
type: string
description: >-
ISO 639-1 language code and ISO 3166-1 region code separated by a
hyphen, e.g. en-US.
example: en-US
model:
type: string
description: Synthesis model to use.
enum:
- simba-english
- simba-multilingual
default: simba-english
audio_format:
type: string
description: Output audio container.
enum:
- wav
- mp3
- ogg
- aac
- pcm
default: wav
options:
$ref: '#/components/schemas/SpeechOptions'
GetStreamRequest:
type: object
required:
- input
- voice_id
properties:
input:
type: string
description: Plain text or SSML to be synthesized to speech.
voice_id:
type: string
description: Id of the voice to be used for synthesizing speech.
language:
type: string
description: >-
ISO 639-1 language code and ISO 3166-1 region code separated by a
hyphen, e.g. en-US.
example: en-US
model:
type: string
enum:
- simba-english
- simba-multilingual
default: simba-english
options:
$ref: '#/components/schemas/SpeechOptions'
SpeechOptions:
type: object
description: Optional synthesis controls.
properties:
loudness_normalization:
type: boolean
description: Whether to normalize the loudness of the output audio.
text_normalization:
type: boolean
description: >-
Whether to apply text normalization (e.g. expanding numbers and
abbreviations) before synthesis.
GetSpeechResponse:
type: object
properties:
audio_data:
type: string
format: byte
description: Base64-encoded synthesized audio.
audio_format:
type: string
description: The format of the returned audio.
enum:
- wav
- mp3
- ogg
- aac
- pcm
billable_characters_count:
type: integer
description: The number of billable characters processed in the request.
speech_marks:
$ref: '#/components/schemas/SpeechMarks'
SpeechMarks:
type: object
description: >-
A nested chunk structure mapping spans of text to start and end times
in the synthesized audio, informing the client when each word is
spoken.
properties:
type:
type: string
example: sentence
start:
type: integer
description: Character offset of the start of the chunk in the input.
end:
type: integer
description: Character offset of the end of the chunk in the input.
start_time:
type: number
description: Start time of the chunk in the audio, in milliseconds.
end_time:
type: number
description: End time of the chunk in the audio, in milliseconds.
value:
type: string
description: The text value of the chunk.
chunks:
type: array
description: Nested speech marks, e.g. words within a sentence.
items:
$ref: '#/components/schemas/SpeechMarks'
Voice:
type: object
properties:
id:
type: string
description: The voice id used as voice_id when synthesizing speech.
display_name:
type: string
description: Human-readable name of the voice.
type:
type: string
description: Whether the voice is a pre-set Speechify voice or a personal clone.
enum:
- shared
- personal
gender:
type: string
enum:
- male
- female
- notSpecified
locale:
type: string
description: Primary locale of the voice, e.g. en-US.
tags:
type: array
items:
type: string
models:
type: array
description: Models the voice supports.
items:
type: string
CreateVoiceRequest:
type: object
required:
- name
- sample
- consent
properties:
name:
type: string
description: Display name for the cloned voice.
sample:
type: string
format: binary
description: A 10-30 second audio sample of the voice to clone.
consent:
type: string
description: >-
JSON string containing the fullName and email of the person who
owns the voice, confirming the voice belongs to you or someone you
represent.
example: '{"fullName":"Jane Doe","email":"jane@example.com"}'
gender:
type: string
enum:
- male
- female
- notSpecified
locale:
type: string
description: Primary locale of the voice, e.g. en-US.