Implementation:Openai Openai python Speech Create

Knowledge Sources	openai-python OpenAI TTS API
Domains	Audio, Speech_Synthesis
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete tool for generating speech audio from text provided by the OpenAI Python SDK.

Description

The Speech resource provides a create() method that synthesizes text into audio using OpenAI's TTS models. It supports multiple voices, output formats (MP3, Opus, AAC, FLAC, WAV, PCM), adjustable speed, and voice style instructions. The response is binary audio data that can be saved to a file or streamed for real-time playback.

Usage

Call client.audio.speech.create() with text input, model, and voice selection. Save the response to a file or use .with_streaming_response for streaming playback.

Code Reference

Source Location

• Repository: openai-python
• File: src/openai/resources/audio/speech.py
• Lines: L48-122 (sync), L145-219 (async)

Signature

class Speech(SyncAPIResource):
    def create(
        self,
        *,
        input: str,
        model: Union[str, SpeechModel],
        voice: Union[str, Literal["alloy", "ash", "ballad", "coral", "echo", "sage", "shimmer", "verse", "marin", "cedar"]],
        instructions: str | NotGiven = NOT_GIVEN,
        response_format: Literal["mp3", "opus", "aac", "flac", "wav", "pcm"] | NotGiven = NOT_GIVEN,
        speed: float | NotGiven = NOT_GIVEN,
    ) -> HttpxBinaryResponseContent:
        """
        Generates audio from text input.

        Args:
            input: Text to synthesize (max 4096 characters).
            model: TTS model ("tts-1", "tts-1-hd").
            voice: Voice selection.
            instructions: Voice style instructions.
            response_format: Output format (default "mp3").
            speed: Playback speed 0.25-4.0 (default 1.0).
        """

Import

from openai import OpenAI
# Access via client.audio.speech.create()

I/O Contract

Inputs

Name	Type	Required	Description
input	str	Yes	Text to synthesize (max 4096 chars)
model	Union[str, SpeechModel]	Yes	TTS model (tts-1, tts-1-hd)
voice	str	Yes	Voice (alloy, ash, ballad, coral, echo, sage, shimmer, verse, marin, cedar)
instructions	str	No	Voice style instructions
response_format	str	No	Output format: mp3, opus, aac, flac, wav, pcm (default mp3)
speed	float	No	Speed 0.25-4.0 (default 1.0)

Outputs

Name	Type	Description
audio	HttpxBinaryResponseContent	Binary audio data (streamable)

Usage Examples

Save to File

from openai import OpenAI
from pathlib import Path

client = OpenAI()
response = client.audio.speech.create(
    model="tts-1-hd",
    voice="alloy",
    input="Today is a wonderful day to build something people love!",
)
response.stream_to_file(Path("output.mp3"))

Streaming Playback

with client.audio.speech.with_streaming_response.create(
    model="tts-1",
    voice="nova",
    input="Streaming audio playback in real time.",
    response_format="pcm",
) as response:
    for chunk in response.iter_bytes(chunk_size=1024):
        play_audio(chunk)

Custom Voice Style

response = client.audio.speech.create(
    model="tts-1-hd",
    voice="shimmer",
    input="Welcome to the future of voice AI.",
    instructions="Speak in a warm, enthusiastic tone.",
    speed=0.9,
    response_format="opus",
)

Related Pages

Implements Principle

• Principle:Openai_Openai_python_Text_to_Speech

Requires Environment

• Environment:Openai_Openai_python_Python_3_9_Plus