OuteTTS API Documentation

Base URL

https://outeai.com/api/v1

Authentication

All API requests require authentication using your access token. Include your access token in the request body under the "token" field.

{
    "token": "your_access_token_here",
    ...
}

TTS Endpoint

POST /tts

Convert text to speech using our TTS models. The endpoint returns a streaming response with generation progress and final audio data.

Request Parameters

Required Parameters

text

String

The text to convert to speech

speaker

Object

Specifies the voice type and speaker name:

Default voices: Use "default": "speaker_name" for built-in voices (e.g., "EN-FEMALE-1-NEURAL")
Custom voices: Use "custom": "speaker_name" for voices you've cloned and stored (using your own named identifier)

Examples:

{"default": "EN-FEMALE-1-NEURAL"} {"custom": "my_clone_001"}

💡 For custom voices, the name must match the identifier you used when creating the speaker.

Optional Parameters

temperature

Float

Sampling temperature (0.1 to 1.0)

Default: 0.4

Response Format

Streaming Response Events

{
    "generation_status": "Warming up TTS model"
}

{
    "generation_status": "Generating linguistic features",
    "text_chunks": 2,
    "current_text_chunk": 1,
    "generated_seconds": 1.5
}

{
    "data": {
        "audio_bytes": "base64_encoded_audio_data",
        "duration": 3.5,
        "request_finished": true
    }
}

Client Library

We provide an official Python client library that handles all the complexity of making API requests and processing streaming responses.

Installation

pip install outeai

Example Usage

from outeai.api.v1 import TTSClient

client = TTSClient(token="your_access_token_here")
output = client.generate(
    text="Hello, how are you doing today?",
    temperature=0.4,
    speaker={"default": "EN-FEMALE-1-NEUTRAL"}
)
# Save the audio
output.save("output.flac")