OpenAI TTS

The OpenAITTSService provides high-quality, natural-sounding voice synthesis using OpenAI's state-of-the-art TTS models. It is seamlessly integrated with the OpenAI ecosystem, making it a perfect choice for agents already leveraging GPT models.

Installation

To use OpenAI TTS, install the required dependencies:

pip install "piopiy-ai[openai]"

Prerequisites

• An OpenAI account and API key (Get yours here).
• Set your API key in your environment:

  export OPENAI_API_KEY="your_api_key_here"

Configuration

OpenAITTSService Parameters

Parameter	Type	Default	Description
api_key	str	None	Your OpenAI API key (defaults to env var).
base_url	str	None	Custom base URL for OpenAI API.
voice	str	"alloy"	Voice ID (alloy, echo, fable, onyx, nova, shimmer).
model	str	"gpt-4o-mini-tts"	TTS model ID.
sample_rate	int	None	Output audio sample rate (OpenAI defaults to 24kHz).
params	InputParams	None	Advanced synthesis controls.

InputParams

Parameter	Type	Default	Description
instructions	str	None	Instructions to guide voice synthesis behavior.
speed	float	None	Voice speed control (0.25 to 4.0).

Usage

Basic Setup

import os
from piopiy.services.openai.tts import OpenAITTSService

tts = OpenAITTSService(
    api_key=os.getenv("OPENAI_API_KEY"),
    voice="nova",
    model="gpt-4o-mini-tts"
)

With Speed Control

from piopiy.services.openai.tts import OpenAITTSService

tts = OpenAITTSService(
    voice="shimmer",
    params=OpenAITTSService.InputParams(
        speed=1.1
    )
)

Notes

• Sample Rate: OpenAI TTS always outputs at 24kHz. Piopiy handles any necessary resampling for your specific transport.
• Latency: The gpt-4o-mini-tts model is highly optimized for performance, providing sub-second TTFB for real-time conversations.
• Streaming: This service uses OpenAI's streaming response to provide audio units as they are generated.